Being right about a problem doesn't mean you're right about your proposed solutions. For the most part, his absolutism limited his ideas to lightly influencing the people trying in earnest to solve those problems rather than actually doing the leg work.* In my old job in academia, I called this "trickle-down public good." He wants to be a symbol for the good and true of free software, but is largely a symbol of ivory tower academic culture throwing shade at people who don't pretend we live in the star trek universe.
* Edit: the problems with software licensing which is the primary thing people know him for. See how many bios even mention his software development work, and then see how many fail to mention his free software zealotry. To even assume his development accomplishments are what people are talking about when discussing his public activities is strange even in a crowd of software developers. If he wrote that same exact software for IBM, or even if he had written it for GNU and not worn crazy outfits to software conventions, nobody would be talking about him 'at all.'
Well now it is. I'm not really sure how you would think that gcc or any other GNU software solved a problem which nobody noticed until twenty years later-- the context for my comment-- but sure.
* Edit: the problems with software licensing which is the primary thing people know him for. See how many bios even mention his software development work, and then see how many fail to mention his free software zealotry. To even assume his development accomplishments are what people are talking about when discussing his public activities is strange even in a crowd of software developers. If he wrote that same exact software for IBM, or even if he had written it for GNU and not worn crazy outfits to software conventions, nobody would be talking about him 'at all.'