There is so much confusion in this thread as to what this is about, I think we should bring out our famously flawed tech analogies. Say the IEEE writes a standard for a networking protocol. It's full of arcane acronyms and technical details, as well as some ambiguities here and there. So then I (as a private person, nothing to do with the IEEE) write a book explaining the protocol in usable terms. I also take some liberties to provide advise on 'best practices' on some of the ambiguous parts. My book is so useful and successful, that even most implementors of the protocol refer to it. Because of that, most implementations behave mostly the same in the corner cases.
Something like this happens quite a lot, to a greater or lesser extent. Nobody (well, copyright abolitionists would, out of principle, but I'm talking people who just want to get stuff done and get on with their life) would argue that I would lose copyright to my book, because it's so successful.
Now the IEEE pays me to write said book. This is where things become murkier (maybe). Because now, has my book become an extension of the standard? Does it depend on the relationship between me and the IEEE? Say, if I have access to drafts of the standard, and I provide feedback, does that make my book closer to being part of the standard? If my book just has clarifications and no opinion in it, does that make a difference? I would still say that, as long as it's possible to implement the standard without the book, it's not 'part' of the standard. But as the standard gets less clear, at some point, it becomes impossible to implement something in an interoperable way without having the book. Should I lose copyright my book in such a case? And does it still matter whether the IEEE has paid for it?
One difference between the IEEE and legislators is that one can say 'screw the IEEE, I'll just make my own protocol'. You can't do that with law. Is that sufficient a difference to make someone lose their copyright? Or should the IEEE just have paid me outright, as a contractor, to write the book, so that they can give it away for free? But then why bother with the book at all? After all, the book is just for making things easier, it's not a 'source' as such, regardless of how much easier it makes implementor's lives.
I don't think this is as clear or as outrageous a case as people are making it out to be here. We don't want to discourage people from making annotations, summaries and explanations. Maybe legislators should just write more clear laws to begin with. But the more you do that, the more you open yourself up to 'backdoors' and 'hole in the law' trickery. Not to mention that it becomes even harder to get political will to support a text in the first place - making perfect the enemy of good enough.