Heh. Now I'm a FSF member, so take what I have to say with a grain of salt.

I really find reading these threads very entertaining. There is a lot of people who focus on personal attacks against Stallman, but that happens in every discussion. Those opinions are similar to anti-Poettering-ism, usually not fact-based and emotionally fuelled, and can be very quickly discarded.

Another large group of people might understand the issue partially, but gets sidetracked somewhere midway through what the mission of the FSF was. To all of these people, I honestly recommend reading up on the history of GNU, and some of Stallman's essays (the Free Software, Free Society collection for example). This same group of people has issues with real world logic, where anything is (according to them) defined as a "love or hate" relationship.

GNU and by extension the FSF were started as an idealistic project. The idea of the whole shebang is to provide "free as in freedom alternative" to proprietary IT solutions - a version of a popular text editor, which is guaranteed to be available forever. A compiler collection, which anyone can use. A de-facto reference C-library implementation, available to anyone. All that under one condition - whatever you do to the code, if you decide to publish the (modified or verbatim) code, it has to be "free as in freedom".

Now, the idea of all that was never to be the one and only option, to burn and erase all other proprietary softwares, and to make sure you can run only the "tagged" stuff. It was actually the exact opposite - to provide you with the (true) freedom of choice, with an alternative. It's like free range vs caged chicken - arguably, there the benefits of each are more clearly cut (conscience vs cost). And it also describes the current goal of FSF - you wouldn't want to buy free range eggs, only to find out that only 95% of them are really free range, and that it's at the distributor's disposal whether he decides to swap some eggs in your package for caged ones...

For a realist / pragmatist, Stallman's point of view is probably far too idealistic to be practicable. What is honourable about RMS is, though, that he eats his cake too - he doesn't use proprietary software (or hardware), he truly drinks water while preaching water. It were the pragmatists who took the GNU bits, the BSD bits and the kernel, and started making distributions. It took long time to have first true "GNU/Linux" distribution, and no one really insists on you using it - it's your choice - but you can't say you're eating organic, when your corn comes from a Monsanto seed.

Stallman and people at FSF usually assume that people reading their statements are not retarded and can perform a critical analysis. I would be very surprised, if in any of their publications you found recommendations on what you "generally should and shouldn't do" (as opposed to "do to remain truly FSF-free"). In any case, I do believe what Stallman says to be incredibly relevant, especially as he is one of the only people taking such a critical stance. I do not pretend I do what he suggests I do, but I do not pretend I am running only FSF-compatible stuff either.