Well, RMS tirelessly advocates for his beliefs, but sometimes his beliefs are regressive and outdated. In particular I remember a debate perhaps 8-10 years ago when the gcc team wanted to expose more info from the compiler back end to enable better support for editor integrations and similar tools, because gcc simply couldn't compete with LLVM in that department. IIRC RMS stepped in with a unilateral decision blocking the effort, saying that it wasn't a big deal anyway because editor integration wasn't an important feature. RMS, a guy who (AFAIK) hasn't done any serious coding in 30 years.
Decisions like this have definitely ended up harming the movement more than helped ultimately. Him refusing to have gcc support that ended up pushing a lot of people to other compilers (llvm) and have sidelined gcc in many spaces.
