I was around at the time that Linux was born. I remember the HURD. I had done some work on Mach while I was in university and I was hugely interested in working on the HURD. I sent email to MIB and essentially received a reply "We don't need you kid". But this is the way it was at the time. People working on GNU were used to doing things in a Cathedral fashion (to put it in ESR's terms). They had small teams where only a few trusted people had input. The rest were users, not collaborators.
The internet was not common at that point and we didn't have things like distributed source repositories. Development was usually done using CVS and you had to be a core contributor to have access. You got access to source code from releases only. And while some projects had fairly regular releases, some did not. For the HURD, I think the idea was that there was no point to having regular releases because it wasn't even self hosting yet.
Linus showed up and changed the world. He said, "Here's this thing I've been working on" and he didn't care who you were. If you sent him a patch, he looked at it. This was completely different from how things used to work for most projects. Linus didn't make you sign over your copyright. He didn't vet you as a core or non-core developer. He just took your patch and evaluated it on its technical merits.
The whole Bazaar approach that is common today, stems from how Linus ran kernel development right from the start. It changed everything in free software development because it was just 100x better. No friction. Welcoming. Not discriminatory.
But if you think RMS is sad about this state of affairs, I think you are terribly mistaken. The HURD was a failure because the world changed under them. Linus came around and showed everybody how software freedom is supposed to work in practice. This took a powerful but small group and expanded its reach to the everyday programmer. RMS's dream of a world where normal developers valued free software over proprietary software came about because of this shift -- and he's smart enough to realise it.
RMS was clumsy about the whole GNU/Linux thing, but Linux only existed because GNU existed. Back in the day, the first thing I would do when I got a new Unix box was to install GNU -- because it was massively better than whatever crap came with the Unix system. Even now, despite the advances of BSD systems, I would still install GNU on top of a BSD kernel if I was running BSD.
Occasionally I hear people asking the question, "If Android is running Linux, why can't I run my Linux apps on my phone?" It drives me crazy. It's because you've got an Android/Linux box instead of a GNU/Linux box. The bit you want is GNU, not Linux. This is precisely why people like RMS wanted to stress the importance of GNU in the equation. I don't agree with his approach, but it's foolish to deny the basis of the argument.
Without RMS's vision, we would not be where we are today. I lived in the world where I had to use $5k per seat proprietary libraries to get anything done. I lived in the world where large corporations who built compilers told me what I was allowed to build. RMS rescued us from that. ESR gave us a vocabulary with which to talk about this stuff. He categorised the kinds of ways people approached things and allowed us to think critically about what we were doing. Linus showed us how to actually make it work. All of these people have disagreed with specific things the others have said, but they also show massive respect for one another -- for good reason. Without them, our lives as developers would be infinitely worse.