I checked 24-bit bitmap rendering today and it has the same set of issues as 16-bit bitmap rendering. The main problem (garbled rendering) is addressed by !8999. Besides that, direct bitmap memory edits are lost on glFinish/glFlush just like they were for 16-bit rendering, although the program in question (Moyo Go Studio) didn't use memory edits so it hasn't been an actual problem thus far. To fix it I think it might be better to use the workaround for 24-bit bitmaps as well. I'm not sure if the workaround would also work for 32-bit bitmaps, since the alpha channel isn't necessarily free for those, although it likely might work. Ideal solution would of course be a compliant software renderer, but that's a lot of work.
I have a test case for 24-bit bitmap rendering on https://gitlab.winehq.org/zowie/wine/-/commit/acc6d8dc9a8fd4f8e2efee4ef1248e... — but that's still only on a draft branch.