I haven't studied the fine print on this service, but I wonder about the (high?) value of knowing a person's music tastes in this level of detail. I have hundreds of playlists on spotify, and I bet you can state a lot of things about me by knowing what music I like.
Last.fm allows you to track your whole listening history (and you can export it or play with their API). I agree it could be interesting to infer something valuable from what a person likes (musically), but I don't know if there's much to read between the lines other than "if user likes artist X probably will like artist Y". You may infer my location (from local artists listened), and from that paired with timestamps you could tell some of my habits (work, school, commuting). Personal taste can hint an age group and gender as well, but it wouldn't be precise at all. Perhaps mood? I don't see much value on that info, compared to what you can gather from navigation history, messages or other services.
Just a tought I've had but what if it will actually reveal more insights about their recommendation and curation algorithms than about your music taste?
This data would be more interesting over time. What new artists have been added to the playlists and have any new songs from already known artists been missed...
