As with SE-0308, I am -1 on this proposal.
This continues to make the language model more complicated by adding special cases. SE-0308 was motivated by rushing in a "quick fix" for critical SwiftUI cases before WWDC. I don't see why we would continue to exacerbate the problems here, instead of investing in a properly scalable fix.
As a procedural issue, is there an implementation of this? The proposal has a 3-year old abandoned patch.
-Chris