FSF's aim is to be the model to follow regarding free software. No one is going to be more radical with free software as they will be, that means that the more they get near proprietary software, the less people will be really serious about only or mostly using free software, because if those that should be %100 libre, are not following their own principles, some tasks would become acceptable to be performed without free software. However, making more usecases to be acceptable through proprietary software, decreases the amount of people interested or interested in the future of free software, basically because you narrowed the free software functioning area, thus, returning again to the situation about "FSF's relationship with [whatever] is harmful to free software users". So what should the FSF do again? allow even more proprietary software? No. What they should do is stay firm, because they are the model to follow and the ones that say how it should be done. And I hope the FSF continues to do what they are doing today, because if they don't, nobody will.

Then, that relationship it is in fact harmful, but it is the only long term way. If you don't like it, you should not blame the FSF, but the vendors, but not really, because there are many that are trying to get near FSF's ideals, and the more radical the FSF is, the closer those vendors/projects will try to reach FSF's level. As on this blog post was mentioned, a company even got to design a phone, with > not one, but two processors just to get FSF's approval. That's amazing in my opinion, not something to be ashamed for. Although I must say, it was such a disgusting trick...