First, is there a clear, immediate, personal benefit for users who tag? If not, then you’re right, users aren’t likely to care. (social use isn’t a very good reason, it needs to be personal)

Second, is the system successful without tagging? That will make sure that the only incentive isn’t to tag…which only leads to SPAM. There needs to be some other benefit.

To your question of bookmarking and categorizing…yes, most implementations of tagging are useful for these reasons. However, they apply very broadly to many industries and types of sites. People need tools to organize!

Like you, I’m anxious to see where tagging can work successfully outside of the initial areas where it has proved successful.

In my current project, we plan to use tagging to supplement our own taxonomic work. It’s been expensive to get sufficient numbers of professionals involved, and our database contains a huge number of objects. So, we’re going to see whether our audience (college and graduate-level students, professors, researchers) will take to it. So at some point, I might have enough stats about those audiences to share.

I have an intranet and I’d love to have the users tag and organize the content as they wish, but we can’t realistically expect the users to do that, now can we?