from the stuff that amazon has published, it seems clear that the copyright holder was involved in telling amazon to stop, since they have the self-service tool to make the book available, the copyright holder could have gone ahead and made it available from that point on. the fact that they choose not to do so seems significant to me.
I would like to see discussion on that larger question, but I felt that I was getting confused over what we were talking about.
with opensource being so common, the probability that _someone_ will publish some code under an opensource license when they don't have the right to is high. once such code is out there with the wrong license terms on it, there is not really any way to 'unpublish' it, there will always be stray copies around with the incorrect license attached, and people will stumble across them and use them, further spreading the problem
it would be a good thing if we had some idea of how this could be dealt with _prior_ to someone ending up in court. I am not saying that the community should in any way support a guilty defendant, but it would make a _lot_ of sense to suggest remedies to the judge in a friend of the court brief that the community can live with rather than risking silly things like orders to shut down sourceforge or google code because they 'repeatedly publish' stolen code (as parts of different projects, all independent from each other, who's authors think they are legitimately using opensource code)