Yet again someone said to me in a meeting: librarians don't like web 2.0, they always push back against it. Ok, so this clearly doesn't describe all of the librarians I hang out with online or any of the ones I work with. My guess is that there are two things that really spawned this. The whole don't-use-wikipedia thing and the whole controlled vocabulary rules thing.
I've described well-meant but overly simplistic heuristics some educators used to teach about evaluating web sites. Along with those, there's typically and outright ban on Wikipedia. The truth is that there is a lot of good and helpful information on Wikipedia, but no, for most things your information search shouldn't end with an encyclopedia. It's no better to copy off of Wikipedia than it is to copy out of World Book. A lot of librarians use Wikipedia and some are even Wikipedians. But we know that it is not sufficient for many information needs.
On the whole controlled vocabulary thing in some circles, I think you're actually getting some computer scientist buy-in on human indexing (or at least annotation) using a controlled vocabulary. The fact is that you get better information retrieval if there's a good controlled vocabulary and some human intervention in the indexing of documents ( and in query formulation). Everyone agrees that this is very expensive, and only scales so far. For real-time things, massive datasets, and in other areas, you don't have this luxury. I don't think librarians begrudge folksonomies, but I know we do get frustrated with U.S.A. not retrieving USA or United States, etc. But we also know tricks for natural language searching and we use them. So if folksonomies are all you have, it's better than nothing, but in many cases it's worth the front loading expense of real indexing for better and faster retrieval and use on the other end.
If you have any doubt about librarian bloggers, catch up in your reading of Walt's work!