That's utter revisionist claptrap.
Stallman's uncompromising stance is pretty evident in the GPL, which is a relatively minor player when compared against more permissive licenses (MIT, Apache, BSD, and -- relevant to the conversation at hand -- MPL). These licenses, by allowing in the "little bit of evil" that is represented by allowing their use in commercial contexts, have been significantly more successful than GPL and similar viral attempts.
You can try to hold him out as a cheerleader in this arena, but in terms of "meeting his philosophical demands," how much of the stuff that Microsoft has released is under viral licenses like GPL?
RMS lost this battle, and it's completely because he won't take compromise of any kind. If the only two options were "closed source or GPL," then the open source movement would have died decades ago. The more compromising stance of organizations of MIT, Berkeley, Apache, and Mozilla -- and the myriad software projects that followed their lead -- is what changed the landscape.