Thanks @xwu for bringing up this point.
The core team took a fresh look at this proposal. We agreed it should go for review based on considering the current implementation of these conditionals. This is similar to what we did with SE-0308, where we did decide to extend the model within the current framework. While moving #if support into the lexer remains an interesting potential direction, no real momentum has come forth on such a direction since the time this proposal was put forth. We thus felt, when we revisited this proposal, that it should be considered by the community on its own merits within the model that currently exists. I should have made this clear when the review was launched.