> Speaking of watchers -- big +1 from me to drop them all at some point. I would start as early as 3.9.

Yeah that was my initial plan, to start the deprecation in 3.9 and finalize the removal in 3.11. We might be able to make an exception for MultiLoopChildWatcher and just remove it from 3.8 though. 

I think it will be necessary to fix the issues in the near future if we want to keep MultiLoopChildWatcher in 3.8 (as they apply to 3.8 and 3.9). I would certainly not be ecstatic about the idea of removing something that was just recently added (similar to how we had to revert the new streaming changes due to API design), but I'm even less in favor of keeping something around that's not stable in a final release. 

Based on my investigation of https://bugs.python.org/issue38323, it seems like it will be a rather complex issue to solve.