FRIENDS OF MCMILLAN PARK, MCMILLAN COALITION FOR SUSTAINABLE AGRICULTURE, and DC FOR REASONABLE DEVELOPMENT v. DISTRICT OF COLUMBIA ZONING COMMISSION and MAYOR'S AGENT FOR HISTORIC PRESERVATION, and VISION MCMILLAN PARTNERS, LLC, Intervenor.

McLeese, Associate Judge: The three orders at issue in these cases ■arise from the efforts of intervenor Vision McMillan Partners, LLC (VMP) to obtain approval to develop a twenty-five-acre parcel of land located on the McMillan Reservoir and Filtration Complex. In the first order, the Zoning Commission approved VMP’s application for a planned unit development (PUD) on the site. In the other two orders, the Mayor’s Agent for Historic Preservation approved permits allowing VMP to demolish certain structures on the site and to subdivide the site. Petitioner Friends of McMillan Park (FOMP) challenges these orders. 1 Specifically, FOMP argues that the project is inconsistent with the District’s Comprehensive Plan and that the Commission failed to adequately explain its conclusions. FOMP also challenges both Mayor’s Agent orders, arguing that the Mayor’s Agent incorrectly determined that the project has “special merit,” incorrectly found that the project’s special merit outweighs the historic-preservation losses that the project would entail, and failed to examine reasonable alternatives *1032 to the project. We vacate the Commission’s order and both Mayor’s Agent orders and remand the cases for further proceedings. L The McMillan Reservoir and Filtration Complex is listed in the D.C. Inventory of Historic Sites and in the National Register of Historic Places. The filtration plant on the site, which used sand to filter drinking water, was constructed in the early 1900s by the U.S. Army Corps of Engineers. The site includes two paved service courts, each with two regulator houses. Cylindrical'portals provide access to twenty subterranean sand-filter beds with vaulted ceilings and supporting arches. Stairs at the corners of the site lead up to a pedestrian path around the perimeter. The landscaping on the site was originally designed by noted landscape architect Frederick Law Olmsted, Jr. The filtration site was decommissioned in 1986, and the federal government sold the parcel of land at issue to the District a year later. The District eventually selected VMP to develop the site. VMP seeks approval to construct a number of buildings as part of the project, including a 115-foot-high health-care facility on the northern portion of the site; a mixed-use building with both a ground-floor supermarket and approximately 280 residential units; 146 individual rowhouses; and a community center. VMP also proposes to create a 6.2-acre park on the southern portion of the site. VMP seeks to demolish all but two of the remaining subterranean sand-filter beds and a number of the portals. VMP also seeks to subdivide the site. VMP proposes to preserve and restore a number of the site’s above-ground resources, including the regulator houses, some portals, and the perimeter path. II. We turn first to the Commission’s order approving the PUD. “We must affirm the Commission’s decision so long as (1) [the Commission] has made findings of fact on each material contested issue; (2) there is substantial evidence in the record to support each finding; and (3) [the Commission’s] conclusions of law follow rationally from those findings.” Howell v. District of Columbia Zoning Comm’n, 97 A.3d 579, 581 (D.C. 2014) (brackets and internal quotation marks omitted). Because the Commission is an expert body, we generally defer to the Commission’s interpretation of the zoning regulations. Id. We will not uphold interpretations that are “plainly erroneous or inconsistent with the regulations.” Citizens Ass’n v. District of Columbia Bd. of Zoning Adjustment, 642 A.2d 125, 128 (D.C. 1994) (internal quotation marks omitted). The PUD process allows the Commission to grant exceptions to otherwise applicable zoning regulations if the PUD offers a “commendable number or quality of public benefits” and “protects and advances the public health, safety, welfare, and convenience.” 11 DCMR § 2400.2 (2016). 2 In deciding whether to approve a PUD, the Commission must weigh “the relative value of the project amenities and public benefits offered, the degree of development incentives requested, and any potential adverse effects.” 11 DCMR § 2403.8 (2016). *1033 The Commission may not approve a PUD that is- inconsistent with the Comprehensive Plan. 11- DCMR § 2400.4; see also D.C. Code § 6-641.02 (2012 Repl.) (amendments to zoning map may not be inconsistent with Comprehensive Plan). The Comprehensive Plan is a “broad framework intended to guide the future land use planning decisions for the District.” Wisconsin-Newarh Neighborhood Coal. v. District, of Columbia Zoning Comm’n, 33 A.3d 382, 394 (D.C. 2011) (internal quotation marks omitted). The Comprehensive Plan includes Area Elements that outline neighborhood-specific development priorities. 10-A DCMR § 104.5, .6 (2016). Another part of the Comprehensive Plan, the Future Land Use Map (FLUM), reflects the District’s policies with respect to future land uses across the city. 10-A DCMR § 225.1 (2016). The FLUM designates residential and commercial areas as being low-density, medium-density, moderate-density, or high-density. 10-A DCMR § 225.2 to .11. The FLUM also includes designations for open space and mixed uses. 10-A DCMR § 225.17, .18. A. Consistency with the Comprehensive Plan FOMP raises several challenges to the Commission’s conclusion that the PUD is not inconsistent with the Comprehensive Plan. First, and most broadly, FOMP argues that the Comprehensive Plan flatly forecloses any high-density development on the site. We disagree. As part of its approval of the PUD, the Commission amended the zoning map and placed the northern part of the site into the C-3-C zoning district. That district is generally applicable to high-density commercial uses. 10-A DCMR § 225.11; 11 DCMR § 105.1 (d)(3)(C) (2016) (describing C-3-C district as “high bulk”). More specifically, the proposed medical building on the northern portion of the site would be 115 feet high and would have a floor-area ratio of 4.08. 3 The proposed height and density of that budding substantially exceed the height and density normally permitted in moderate- or medium-density commercial districts such as C-2-A, C-2B, and C-3-A. See 11 DCMR §§ 770.1, 770.6, 771.2 (2016) (describing maximum building height and density in C-2-A, C-2-B, and C-3-A districts); 10-A DCMR § 225.9, .10 (describing C-2-A, C-2-B, and C-3-A districts as moderate- or medium-density zones). Even taking into account the additional flexibility available through the PUD process, the proposed floor-area ratio would exceed that permitted in C-2-A, C-2-B, and C-3-A districts. See 11 DCMR § 2405.2, .3 (2016) (describing maximum floor-area ratio permissible for PUD in C-2-A, C-2-B, and C-3-A districts). The Commission thus correctly acknowledged that the PUD contemplates some “high-density” development on the site. 4 *1034 As FOMP points out, the FLUM designates future uses at the McMillan site as “moderate density commercial,” “medium density residential,” and “parks, recreation, and open space.” We agree with the Commission, however, that permitting some high-density development on the site does not necessarily make the PUD inconsistent with the FLUM. The FLUM explicitly contemplates two ways in which more intensive development than is otherwise reflected in the FLÚM may be permissible: (1) a larger development that as a whole is consistent with the FLUM designation may contain individual buildings with greater height or density; and (2) the PUD process may permit greater height or density. 10-A DCMR § 226.1 (c) (2016). Here the Commission concluded that, when the entire site is taken into account, the PUD’s overall density is consistent with that permitted in moderate-density commercial zones. We do not understand FOMP to dispute that conclusion. The Commission thus reasonably determined that the PUD as a whole was not inconsistent with the FLUM. FOMP also points out that the Mid-City Area Element states that development on the McMillan site “should consist of moderate- to medium-density housing, retail, and other compatible uses.” 10-A DCMR § 2016.9 (2016). We agree with FOMP that the high-density use approved in the PUD is not consistent with that policy. Unlike the FLUM designation discussed above, the Mid-City Area Element does not appear to contemplate any high-density uses on the site. We have emphasized, however, that “even if a proposal conflicts with one or more individual policies associated with the Comprehensive Plan, this does not, in and of itself, pre-clude the Commission from concluding that the action would be consistent with the Comprehensive Plan as a whole.” Durant v. District of Columbia Zoning Comm’n (Durant I), 65 A.3d 1161, 1168 (D.C, 2013). The Comprehensive Plan reflects numerous “occasionally competing policies and goals,”'and, “[e]xcept where specifically provided, the Plan is not binding.” Id. at 1167, 1168 (internal quotation marks omitted). Thus “the Commission may balance competing priorities” in determining whether a PUD is consistent with the Comprehensive Plan as a whole. D.C. Library Renaissance Project/West End Library Advisory Grp. v. District of Columbia Zoning Comm’n, 73 A.3d 107, 126 (D.C. 2013). FOMP argues that the specific language of the Mid-City Area Element is mandatory and necessarily prevails over other more general policies reflected in the Comprehensive Plan. We conclude to the contrary. The Mid-City Area Element’s policy favoring moderate- and medium-density development on the site is not expressed in unambiguously mandatory terms. Rather, that policy is one of several “basic objectives [that] should be pursued” in developing the site, and the policy states that development on the site “should consist of moderate- to medium-density housing, retail, and other compatible uses.” 10-A DCMR § 2016.4, .9 (emphasis added). The term “should” often is properly interpreted to “suggest[ ] or recommend! ] a course of action,” rather than to “describe! ] a course of action that is mandatory,” United States v. Maria, 186 F.3d 66, 70 (2d Cir. 1999) (describing former interpretation as “the common meaning” of “should”). *1035 The Commission thus reasonably-concluded-that the Comprehensive Plan does not flatly prohibit any high-density development on the site. We' emphasize, however, that the Comprehensive Elan’s provisions have substantial force even if they are not mandatory. The policies reflected in the Comprehensive Plan are intended to “[g]uide executive and legislative decisions on matters affecting the District and its citizens.” D.C. Code § 1-306.01 (b)(2) (2012 Repl). The Commission cannot simply disregard some provisions of the Comprehensive Plan on the ground that a PUD is consistent with or supported by other- provisions of the Comprehensive Plan. Rather, if the Commission approves a PUD that is inconsistent with one or more policies reflected in the Comprehensive Plan, the Commission “must recognize these policies and explain [why] they are outweighed by other, competing considerations _” Durant I, 65 A.3d at 1170. FOMP argues-that the Commission failed to adequately explain why it was necessary to disregard the policy favoring medium- and moderate-density development on the site in order to advance other competing policies reflected in the Comprehensive Plan. We agree. The Commission stated that permitting high-density development on the'northern portion of the site was “a critical and essential part of fulfilling the parks, recreation, and open space designation of the [FLUM], while at the same time achieving other elements of the Comprehensive Plan and the city’s strategic economic plan.” FOMP argued before the Commission, however, that the other policies reflected in the Comprehensive Plan could be advanced even if development on the site were limited to medium- and moderate-density uses. The Commission neither provided a specific basis for concluding to the contrary nor stated • reasons for giving greater weight to some policies than to others. We therefore vacate the Commission’s order and remand for further proceedings. See, e.g., Durant v. District of Columbia Zoning Comm’n (Durant II), 99 A.3d 253, 262 (D.C. 2014) (vacating Commission’s order approving PUD and remanding for further proceedings, because “the Commission has not explained why the various policies at issue conflict so as to require a trade-off among them”). Our “remand is not solely for the purpose of redrafting findings and conclusions to facilitate pur review and - reinforce the [Commission’s] decision. The [Commission] may conduct further hearings or even reach a different result.” Ait-Ghezala, 148 A.3d at 1218, 2016 WL 6659496 at *5 (ellipses and internal quotation marks omitted). In' a related point, FOMP contends that the Commission failed to adequately address a. number of provisions in the Comprehensive Plan that FOMP argues weigh against approval of the PUD, including provisions discouraging the placement of large buildings near low-density residential neighborhoods, 10-A DCMR §§ 305.11, 309.10, 309.15 (2016), and a provision encouraging geographic dispersion of healthcare facilities, 10-A DCMR § 1105,1 (2016). We agree that such provisions merit explicit consideration on remand, B. Other Objections to the Commission’s Order Although we have already concluded that the Commission’s order must be vacated, we briefly address several additional issues that could affect proceedings on remand. 1. Preservation of Open Space FOMP asserts that the Mid-City Area Element requires preservation of open space on the site. It is true that the Mid- *1036 City Area Element provision relating to open space on the site uses the word “require.” 10-A DCMR § 2016.5 (“Require that reuse plans for the McMillan Reservoir Sand Filtration site dedicate a substantial contiguous portion of the site for recreation and open space”). That provision, however, appears in a larger framework that describes the site-specific provisions in a less mandatory way—as “basic objectives [that] should be pursued in the re-use of the McMillan Sand Filtration site.” 10-A DCMR § 2016.4 (emphasis added). Moreover, even Comprehensive Plan policies that are expressed in entirely mandatory terms may conflict with each other. In such circumstances, the Commission would need to determine which policy to pursue. For these reasons, we are doubtful that the policy favoring retention of open space would be mandatory in all circumstances. In any event, we do not agree with FOMP’s argument that the need to preserve open space could never be used to justify the inclusion of high-density development on the site. For example, if including some high-density development on the site were the only feasible way to retain a substantial part of the property as open space and make the site usable for recreational purposes, then the Commission might be able to permissibly conclude that the need to preserve open space justified the inclusion of some high-density development on the site. 2. Adverse Impacts FOMP argues that the Commission failed to adequately address a variety of asserted adverse impacts of the PUD, including environmental problems, destabilization of land values and displacement of neighboring residents, and increased demand for essential public services. In a number of respects, we agree. We turn first to the PUD’s impact on the environment. The Comprehensive Plan contains an element directed to the potential environmental effects of development. 10-A DCMR §§ 600-630 (2016). In addition, the PUD regulations (1) specifically direct the Commission to consider the environmental benefits associated with a PUD, 11 DCMR § 2403.9 (h); and (2) generally direct the Commission to consider “any potential adverse effects” associated with a PUD, 11 DCMR § 2403.8. These provisions indicate that the Commission must consider environmental impacts, both in deciding whether a PUD is consistent with the Comprehensive Plan and in deciding whether a PUD would have adverse effects. The Commission in this case did consider environmental impacts to a degree. It specifically referred to evidence regarding water and sewer management, low-impact design techniques, and LEED certification for the buildings on the site. The Commission also stated that “the Applicant is proposing sufficient public benefits that outweigh environmental impacts.” The basis for the Commission’s statement is not clear, however. FOMP raised a number of environmental concerns, including claims that the PUD would increase pollution, noise, waste, emissions, and use of water, electricity, and gas. The Commission declined to address those concerns, stating that “[ejnvironmental studies are best conducted by the District Department [of] the Environment ... and will be part of the building permit process.” In declining to fully address FOMP’s environmental concerns, the Commission relied upon Foggy Bottom Ass’n v. District of Columbia Zoning Comm’n, 979 A.2d 1160 (D.C. 2009). We do not understand Foggy Bottom to permit the Commission to decline to consider environmental impacts when reviewing a PUD application. *1037 The issue in Foggy Bottom was whether the Commission was required to delay consideration of a PUD application until ah environmental-impact statement had been prepared. Id. at 1163. After carefully examining the pertinent statutory provisions, we concluded that the Commission was not required to wait for an environmental-impact statement. Id. at 1164-67. In this case, FOMP does not contend that the Commission must delay consideration of the PUD until completion of an environmental-impact statement. Rather, FOMP contends that the Commission has a clear responsibility under the applicable statutes and regulations to assess environmental impacts when deciding whether to grant a PUD application. For the reasons already stated, we agree. Cf., e.g., Levy v. District of Columbia Bd. of Zoning Adjustment, 570 A.2d 739, 750-52 (D.C. 1990) (Board of Zoning Adjustment erred by declining to consider certain concerns about proposed development on ground that Mayor and other agencies had authority to address those concerns). 5 Second, FOMP argues that the Commission failed to adequately consider the PUD’s potential effects on neighboring property values and the risk that neighborhood residents would be displaced. The Comprehensive Plan specifically addresses the topics of property values and displacement. E.g., 10-A DCMR §§ 205.6, 218.1, 218.3, 508.1, 2502.5 (2016). The Commission therefore must appropriately address those topics when deciding whether a PUD is consistent with the Comprehensive Plan and whether a PUD would have adverse effects. The Commission acknowledged FOMP’s concerns that the PUD would accelerate gentrification, increase land values, and result in a net loss of affordable housing. The Commission nevertheless dismissed those concerns as conclusory and unsupported by evidence, thus apparently placing the burden on FOMP to prove a potential adverse effect. As FOMP points out, however, the PUD regulations state that “[t]he applicant shall have the burden of proof to justify the granting of the application .'...” 11 DCMR § 2403.2; see also 11 DCMR § 2407.6 (2016) (“At the public hearing, the applicant shall carry the burden of justifying the proposal.”); Cathedral Park Condo. Comm. v. District of Columbia Zoning Comm’n, 743 A.2d 1231, 1246-47 (D.C. 2000) (PUD applicant has burden of proof). Moreover, the Commission may not approve á PUD unless it finds that the PUD “protects and advances the public health, safety, welfare, and convenience.” 11 DCMR § 2400.2; see also 11 DCMR § 2403.8 (in deciding PUD application, Commission must weigh “the relative value of the project amenities and public benefits offered, the degree of development incentives requested, and any potential adverse effects”). It is unclear from these provisions that the Commission permissibly required FOMP to bear thé burden of proving that the PUD would give rise to adverse effects. On remand, the Commission thus must either place the burden of proof on YMP or explain why a different allocation is permissible under the PUD regulations. VMP suggests that the Commission- did adequately address the-“neighborhood impact” of the PUD. In support of this suggestion, VMP points to testimony from *1038 Advisory Neighborhood Commission members addressing various issues. The mere existence of testimony touching on a topic, however, does not demonstrate that the Commission considered and adequately addressed that topic. On remand, the Commission should explicitly address FOMP’s arguments concerning issues of gentrification, land values, and displacement. Third, FOMP argues that the Commission did not adequately address whether the PUD would place an undue strain on public services. See generally, e.g., D.C. Code § 6-641.02 (zoning regulations shall be designed to create conditions favorable to efficient public services); 10-A DCMR §§ 1100-14 (2016) (Comprehensive Plan policies relating to provision of public services); 11 DCMR § 2403.3 (PUD may not have unacceptable impact on operation of city services). FOMP attributes this problem in part to the fact that the pertinent D.C. agencies, such as the Department of Housing and Community Development, the Metropolitan Police Department, and the Fire and Emergency Medical Services Department, failed to submit written reports in response to the PUD application. See 11 DCMR §§ 2407.3, 2408.4 (2016) (Commission must, in processing PUD applications, submit applications to D.C. Office of Planning to prepare assessment that “shall include reports in writing from all relevant District agencies and departments, including, but not limited to, the Departments of Transportation and Housing and Community Development — ”). It appears that a number of relevant District agencies were invited to provide written reports concerning the PUD but did not do so. It also appears that, with the exception of a discussion of traffic impacts, the Commission’s order did not address whether the PUD would place an undue strain on public services. It is not clear whether FOMP squarely presented this concern to the Commission, but VMP does not argue that the issue is not properly before us. VMP does argue that the Commission was not required to obtain written statements from relevant agencies. Specifically, VMP asserts that the Commission was only required to solicit comments from those agencies through the Office of Planning, and that in any event the Commission had a report from the Department of Transportation and a 2002 report from the Department of Housing and Community Development. We leave it for the Commission to address these issues on remand. III. •[9] We also vacate and remand the Mayor’s Agent’s orders. Under the Historic Landmark and Historic District Protection Act (the “Preservation Act”), the May- or’s Agent may issue a permit to demolish or subdivide -a historic landmark if the planned demolition or subdivision is “necessary in the public interest.” D.C. Code §§ 6-1104 (a), (e); 6-1106 (a), (e) (2016 Supp.). Demolition and subdivision are “[n]ecessary in the public interest” if they are “necessary to allow the construction of a project of special merit.” D.C. Code § 6-1102 (10) (2016 Supp.). A project has special merit if it provides “significant benefits to the District of Columbia or- to the community by virtue of exemplary architecture, specific features of land planning, or social or other benefits having a high priority for community services.” D.C. Code § 6-1102 (11). If a project has special merit, the Mayor’s Agent must balance that -special merit against the harm to historic-preservation values that would result from the demolition or subdivision. Citizens Comm, to Save Historic Rhodes Tavern v. District of. Columbia Dep’t of Hous. & Cmty. Den, 432 A.2d 710, 715-16 (D.C. 1981). *1039 Our review of a Mayor’s Agent’s decision is “limited and narrow.” Embassy Real Estate Holdings, LLC v. District of Columbia Mayor’s Agent for Historic Pres,, 944 A.2d 1036, 1050 (D.C. 2008) (internal quotation marks omitted). “We must uphold the Mayor’s Agent’s decision if the findings of fact are supported by substantial evidence in the record considered as a whole and the conclusions of law flow rationally from these findings.” Kalorama Heights Ltd. P’ship v. District of Columbia Dep’t of Consumer & Regulatory Affairs, 655 A.2d 865, 868 (D.C. 1995). When the Mayor’s Agent’s “decision is based on an interpretation of the statute and regulations [the Mayor’s Agent] administers, that interpretation will be sustained unless shown to be unreasonable or in contravention of the language or legislative history of the statute.” Id. (internal quotation marks omitted). A. Special Merit We turn first to the Mayor’s Agent’s determination that the project has special merit. “[A] proposed amenity [must] meet a high standard in order to qualify as a ‘special merit’ project, the construction of which would warrant demolition of a building of historical significance.” Committee of 100 on the Fed. City v. District of Columbia Dep’t of Consumer & Regulatory, Affairs, 571 A.2d 195, 200 (D.C. 1990). “[F]actors which are common to all projects are not considered as special merits.” Id. (internal quotation marks omitted). FOMP does not appear to dispute, and we therefore take as a given, that the project has at least some special merit because the project includes the construction of affordable housing beyond what is legally required. FOMP does, however, challenge other aspects of the Mayor’s Agent’s conclusion that the project has special merit. FOMP argues that features of a project that do not rise to the level of “special merit” when considered in isolation cannot contribute to the special merit of the project. We see no basis in the applicable statutes or regulations to foreclose the possibility that a project’s special merit could rest in whole or in part on a combination of features that in isolation would not necessarily rise to the level of special merit. To the contrary, the Preservation Act refers in the plural to “specific features of land planning,” D.C. Code § 6-1102 (11), which suggests that special merit can arise from the combination of more than one land-planning feature. Cf. Citizens Comm., 432 A.2d at 717 n.13 (describing projected economic benefit to city as “another factor militating in favor of a finding of special merit”). On the other hand, we agree with FOMP that the Mayor’s Agent’s orders do not explain with sufficient clarity which “specific features of land planning” the Mayor’s Agent relied upon and why those features combined to support a conclusion of special merit. We turn first to the Mayor’s Agent’s statement that “the totality of the plan ... created the special merit.” We have emphasized that special merit is a “high standard” and that a conclusion of special merit cannot rest on benefits common to all projects. Committee of 100, 571 A.2d at 200. It therefore is critical that the Mayor’s Agent precisely and clearly identifies the specific features of land planning on which the Mayor’s Agent relies to support a conclusion of special merit. The Mayor’s Agent also must specifically explain why those features are “sufficiently special” as to rise to the level of special merit. Id. (internal quotation marks omitted). A broad focus on the overall benefits flowing from a pro- *1040 jeet runs beyond the task assigned to the Mayor’s Agent. Cf. District of Columbia Pres. League v. Dep’t of Consumer & Regulatory Affairs, 646 A.2d 984, 990 (D.C. 1994) (“There is nothing in the Preservation Act that allows the Mayor’s [A]gent to engage in a balancing of interests which takes into account such factors as the cost of refurbishing the dilapidated structure and the threat it poses to the safety and welfare of the community. On the contrary, the limited task of the Mayor’s [A]gent is to evaluate a demolition application in accordance with the Preservation Act, and nothing more.”). Moreover, if the special-merit inquiry could appropriately focus on the “totality” of the benefits arising from a project, then presumably the Mayor’s Agent should also take into account all of the project’s adverse impacts. Under such an approach, the Mayor’s Agent would function essentially as a second Zoning Commission, evaluating all of the benefits and adverse impacts associated with projects requiring a permit from the Mayor’s Agent. We conclude that the Preservation Act assigns the Mayor’s Agent the more discrete role of determining whether one or more specific attributes of a project, considered in isolation or in combination, ,rise to the level of special .merit, thus triggering a balancing of those special-merit benefits against historic-preservation losses. See D.C. Code §§ 6-1102 (11), 1104 (e), 1106 (e); Committee of 100, 571 A.2d at 200; Citizens Comm., 432 A.2d at 715-16. Second, FOMP takes issue with the Mayor’s Agent’s statement that “[c]on-sistency with the Comprehensive Plan may help provide the basis for a project’s special merit.” That statement is potentially confusing. As FOMP points out, overall consistency with the Comprehensive Plan is a legal requirement for PUD approval and zoning amendments. 11 DCMR §§ 2400.4, 2403.4; D.C. Code § 6-641.02. Moreover, the fact that a project does not run afoul of the Comprehensive Plan as a whole does not necessarily demonstrate anything about whether the project is beneficial, much less whether the project has special .merit. Specific provisions in the Comprehensive Plan, however, can play a key role in the special-merit inquiry. Such provisions can, for example, support a conclusion that particular features of land planning are of sufficient significance as to rise to the level of special merit. See Edwin L. Fountain & M. Jesse Carlson, The “Special Merit” Provision for Demolition or Alteration of Historic Properties Under the District of Columbia Historic Preservation Act, SJ053 ALI-ABA 531, 539-40 (2004) (“The more an applicant can tie elements of the proposed project to specific preferred land uses set out in the D.C. Comprehensive Plan, the more likely it is that the Mayor’s Agent will approve the project under this element of special merit. However, mere compliance with applicable zoning requirements is not enough to establish ‘special features of land planning.’”); cf. Committee of 100, 571 A.2d at 201-02 (discussing whether specific provisions of Comprehensive Plan supported conclusion that project had special merit). Such policies must be specifically identified, and the Mayor’s Agent must explain why those policies are “sufficiently special” as to support a conclusion of special merit. Committee of 100, 571 A.2d at 200. With respect to at least some of the features identified in the Mayor’s Agent’s orders in this case, such as the fact that the project involves “a mix of market and subsidized residential units and needed retail,” the Mayor’s Agent has not yet provided such an explanation. Third, the Mayor’s Agent at one point says that “the medical offices themselves do not contribute to the special merit of the project.” Elsewhere, however, the *1041 Mayor’s Agent appears to rest the finding of special merit in part on the conclusion that the project provides “an office use well adapted to the location^ which is] adjacent to” nearby hospitals. On remand, the Mayor’s Agent should clarify this point. Fourth, FOMP argues that the Mayor’s Agent should not have considered the inclusion of a park on the southern portion of the site and the restoration of certain structures on the site as features contributing to the special merit of the project. We agree. It is true that specific provisions of the Comprehensive Plan highlight the importance of preserving open space and physical resources on the site. 10-A DCMR § 2016.5, .6. Nevertheless, the project’s historic-preservation benefits are appropriately treated as reducing the project’s net historic-preservation loss rather than as contributing to the project’s special merit. In this case, the Mayor’s Agent considered historic-preservation benefits both as contributing to the proj'ect’s special merit and as reducing the overall preservation losses that the project would entail. Considering historic-preservation benefits at both steps of the analysis poses a risk of double-counting. Moreover, that a project has some- historic-preservation benefits that help to offset the project’s historic-preservation losses does not logically provide a basis upon which to conclude that the project-provides a “significant benefit” that- rises to the level of special merit and that would justify demolition or subdivision of a historic landmark. In contrast, if a project on balance benefits historical-preservation interests more than it harms those interests, the Mayor’s Agent need not make a special-merit finding before approving demolition or subdivision. See D.C. Code §§ 6-1102 (10), 1101 (b), 1104 (e), 1106 (e) (2012 Repl.); District of Columbia Pres. League v. District of Columbia Dep’t of Consumer & Regulatory Affairs, 711 A.2d 1273, 1275 (D.C. 1998). For these reasons, we vacate the May- or’s Agent’s orders and remand for further proceedings. As with the remand to the Zoning Commission, the remand to the Mayor’s Agent “is not solely for the purpose of redrafting findings and conclusions to facilitate our review and reinforce the [Mayor’s Agent’s] decision. The [Mayor’s Agent] may conduct further hearings or even reach a different result.” Ait-Gheza-la, 148 A.3d at 1218, 2016 WL 6659496 at *5 (ellipses and internal quotation marks omitted). 6 B. Balancing of Special Merit and Historic-Preservation Loss If a project has special merit, the Mayor’s Agent must “balance the historical value of the particular landmark against the special merit of the proposed project.” Citizens Comm., 432 A.2d at 716. Given the need for farther consideration of the question of the project’s special merit, the Mayor’s Agent will need to reconsider the balancing of special merit against historic-preservation loss. We briefly address two additional points that could affect that balancing. First, as we have already explained, the Mayor’s Agent’s task is not to balance all of the benefits of the project against all of the adverse impacts of the project. That broader task is assigned to the Zoning Commission. Rather, the May- or’s Agent’s task is to balance the special *1042 merit of the project—the specific aspects of the project that provide “sufficiently special” benefits—against one particular adverse impact—the net historic-preservation loss that the project would entail. Committee of 100, 571 A.2d at 200. Second, the Mayor’s Agent noted that VMP “equivocate[d]” about whether the project as constructed would' actually preserve some of the historic structures on the site. The Mayor’s Agent addressed this issue by requiring that VMP obtain the Historic Preservation Review Board’s approval for any decision not to retain those structures. FOMP argues that the Mayor’s Agent was not permitted to leave the amount of historic-preservation loss unsettled and to the discretion of another decision-maker, We agree. Cf Committee of 100, 571 A.2d at 204-05 (remanding for further proceedings where Mayor’s Agent relied on future recordation of covenant to ensure actual implementation of special-merit features of project, because “further proceedings [were] needed to flesh out the nature of the covenant .., before the Mayor’s Agent [made] her final decision on the demolition application”). C. Necessity Finally, the Mayor’s Agent was required to determine whether the proposed demolition and subdivision were necessary to allow the construction of a project of special merit. D.C. Code §§ 6-1102 (10), 1104 (e), 1106 (e). We briefly address one issue that could affect further proceedings on remand with respect to that determination. FOMP argues that VMP was required to demonstrate that demolition and subdivision were necessary to obtain the project’s special-merit benefits, rather than that demolition and subdivision were necessary to construct the particular project proposed by VMP. FOMP further argues that VMP failed to make the required showing and that, to the contrary, the asserted special-merit benefits could be achieved through less-intensive development that would permit greater preservation of the historical value of the site. In rejecting these arguments, the Mayor’s Agent acknowledged that demolition would not be necessary if “minor modifications” to the project could avoid or minimize the need for demolition. With that exception, however, the Mayor’s Agent’s stated that the necessity inquiry turns on whether demolition or subdivision would be “necessary to construct [this] project, not one entirely different” or “substantially different.” We agree with FOMP that the Mayor’s Agent erred in framing the necessity inquiry in this way. The Preservation Act protects historic landmarks by requiring a special showing before they may be demolished or subdivided, Kalorama Heights, 655 A.2d at 873-74. Among other things, an applicant seeking approval to demolish or subdivide a- historic landmark bears the burden of showing that demolition or subdivision is “necessary.” D.C. Code.§§ 6-1102(10); 6-1104 (e), (f); 6-1106 (e), (f); cf., e.g., Kalorama Heights, 655 A.2d at 869 (“The applicant has the burden of proving entitlement to a demolition permit. In meeting this burden, the applicant must show that it considered alternatives to-the total demolition of the historic building and that these alternatives were not reasonable.”) (citation omitted). Although an applicant need not demonstrate that there are no other feasible alternatives, an applicant “should be required to • show that all reasonable alternatives were considered.” Citizens Comm., 432 A.2d at 718. “Reasonableness must be imputed into the <neeess[it]y’ standard .... [F]actors including but not limited to cost, delay, and technical feasibility become proper considerations' for determining ‘necess[it]y\ Each of these factors has bearing on whether there are viable *1043 alternatives to demolition available, and the answer to this question determines necessity.” Id, (ellipses and internal quotation marks omitted). If a reasonable alternative would achieve the same special-merit benefits of a project while avoiding or reducing the need for demolition or subdivision, thereby reducing the adverse impact on historic-preservation interests, then the Mayor’s Agent cannot properly conclude that the proposed demolition or subdivision is “necessary to allow the construction of a project of special merit.” D.C. Code § 6-1102 (10). That is true without regard to whether the reasonable alternative would entail “substantial” or instead only “minor” changes to the project as proposed. There is no basis in the language or purposes of the Preservation Act for trying to draw a line between “minor” and, “substantial” changes, and attempts to draw such a line would prove difficult. For the foregoing reasons, we vacate the Commission’s order and the Mayor’s Agent’s two orders and- remand these cases for further proceedings. So ordered. Footnotes [1] Two other associations—McMillan Coalition for Sustainable Agriculture and DC for Reasonable Development—also petitioned for review of the Zoning Commission’s order. Although VMP argues that these associations lack standing to challenge the Commission's order, VMP does not dispute FOMP’s stand-tag. Because FOMP has standing and has adopted all of the arguments made by the other two associations, we need not decide whether the other two associations have standing. E.g., Sahrapour v. LesRon, LLC, 119 A.3d 70,4, 707 n.1 (D.C. 2015). [2] , The Zoning Commission promulgated new zoning regulations effective September 6, 2016. 11-A DCMR § 100.3 (2016). Those regulations are not applicable to this proceeding. 11-A DCMR § 100.4 (b); Ait-Ghezala v. Dis trict of Columbia Bd. of Zoning Adjustment, 148 A.3d 1211, 1214-15, 2016 WL 6659496, at *2 n.2 (D.C. 2016). All citations to the zoning regulations in this opinión refer to the prior regulations. [3] ‘‘[The floor-to-area ratio] is a measure of building density arid is determined by dividing the gross floor area of all buildings ori a lot by the area of that lot.” Durant v. District of Columbia Zoning Comm’n (Durant III), 139 A.3d 880, 882 (D.C. 2016) (internal quotation marks omitted). [4] VMP suggests that the C-3-C district does not necessarily correspond only to high-density commercial uses, because the provisions categorizing certain zoning districts as consistent with moderate- and medium-density commercial uses state that "other districts may apply.” 10-A DCMR § 225.9, .10. The Commission did not rely on that rationale, instead acknowledging that the PUD proposed high-density development of the northern portion of the site. In any event, we do not view the references to the possibility that other districts might apply as supporting a conclusion that buildings permissible only in a C-3-C district could reasonably be viewed as medium- or moderate-density uses. Cf. Durant III, 139 A.3d at 884 (although higher- *1034 density buildings may be permissible in moderate-density areas, that "does not mean that such buildings are themselves necessarily understood to be moderate-density in character"). [5] We express no view about whether, and if so in what circumstances, the Commission may appropriately defer to the prior conclusions of other expert agencies. Cf., e.g., D.C. Library Renaissance Project, 73 A.3d at 121 (‘‘Other courts have held that agencies in some circumstances appropriately can, or even must, defer to the prior determination of another agency with overlapping authority.”). [6] FOMP and amici raise concerns about the Mayor's Agent’s consideration of VMP’s demolition and subdivision applications in separate proceedings. We assume that the Mayor’s Agent will address the demolition and subdivision applications together and in a comprehensive manner on remand. We therefore see no need to further address this issue.