I may be biased in that I have not written much C-code in years (decades?), but whenever I find a codebase covered in IFDEFs, I start assuming that every single new IFDEF introduces a new condition into the system which has not been properly nor recently tested, and that the software is horribly broken.

For me, fixing software, is just as often removing IFDEFs together with unmaintained and broken code. It's IMO much better to be honest about "this isn't supported", than pretend you support something you don't.

And this seems to be another one of those stories, coupled with a (bad?) case of NIH.