But if Stallman is actually pursuing the OP's idea of extending the Overton window, then he's specifically not trying to accomplish short-term goals. Other people can and should be working on those short-term goals. Stallman, meanwhile, is trying to stretch the playing field out far enough that the smaller goals seem very reasonable by comparison to the extreme that he creates.

I'm not sure that Stallman is intentionally positioning himself this way (he seems to sincerely believe his all-or-nothing stance), but I could see it actually working. The big problem with Stallman is that he has to be at least moderately credible in order to extend the Overton window, but he seems to perpetually be right on the edge of being completely dismissed. If that happens then the window snaps back shut.