I am not yet convinced that https://github.com/reactos/reactos/pull/2760 was actually a correct fix.
I just realized that 0.4.14-RC-106-g6702f8a is actually affected by the glitch that this PR once aimed to fix.
The sndrec32 of 0.4.14-RC-106-g6702f8a shows this glitch only when run on ros.
The sndrec32 of 0.4.14-RC-106-g6702f8a does not show this glitch when run on XPSP3
This was the first thing that made me suspecting the patch targetting sndrec might have been wrong, or at least not all that was needed.
Moreover the fact that ancient 0.4.7-release-92-gd339297_wasStillFine.png was rendering the background of those static controls still fine, is a pretty clear indicator that we actually had a regression here somewhen between 0.4.8-dev'-1 and 0.4.14-dev-1562-g4d2d2db, which made the rendering wrong.
We should bisect this to the guilty rev to be able to provide a better fix, in deeper parts of ros (e.g. win32ss) as well, instead of just fixing the sndrec32-sources.
Unfortunately the user 'ghost' who provided the initial fix in sndrec32 by https://github.com/reactos/reactos/pull/2760, got deleted from github meanwhile and can therefore not be contacted easily.
You will not be able to see this on master head, because the underlying tiny incompatibility in Win32SS got hidden again by the sndrec32-fix in 0.4.15-dev-190-gd839e3d.