What advantages do we gain from this change?

We gain way less risk of getting burnt by the next VS minor update silently breaking support for an unsupported Windows version. They cannot break 21H2 anytime soon as that version is an LTSC one.

Also note that we only have 3.36% of Windows 10/11 users using Education or Enterprise editions:

 1b_PRODUCT_ENTERPRISE_N                            0.03% #
 79_PRODUCT_EDUCATION                               1.44% ##
 7a_PRODUCT_EDUCATION_N                             0.02% #
 7d_PRODUCT_ENTERPRISE_S                            1.27% ##
 7e_PRODUCT_ENTERPRISE_S_N                          0.03% #
 a4_PRODUCT_PRO_FOR_EDUCATION                       0.32% #
 a5_PRODUCT_PRO_FOR_EDUCATION_N                     0.04% #
 bc_PRODUCT_IOTENTERPRISE                           0.02% #
 bf_PRODUCT_IOTENTERPRISE_S                         0.19% #

, which implies at least half of the Win10 20H2 and Win10 21H1 users are already using an EOL Windows version right now.

If we release at easter and not bump the requirements before that, all we keep is 4 weeks of support for Education/Enterprise Win10 20H2, and nothing else.

... and the only real feature differences between supported standard build and supported legacy build is improved time zone correctness and arm support.

Then I guess we can go ahead.

r18524 / 1.31.00.21 / 0.7.0-pre.16