For me, this is one of the best Web2.0 overview diagram. (with all due respect to Tim O'Rielly's meme diagram, which did get this all started). The single most important thing is the flow, and the realization of the differing categorizations of data sources, both trusted and non. The reason I liked this more than most of the Web2.0 diagrams is that it is the first to challenge my thinking and assumptions.
Specifically, look at the example sites of UserData - CraigsList and Wikipedia are both on the list. At first, I thought, "but wait, those are not user data" - and even went as far as to think there should be another source more clearly linked to "user" data. But I am wrong, and this is where Peter's diagram was inspiring.
