Former Grooveshark engineer here. Somehow I hadn't heard about this project until now. In case the author reads this - feel free to reach out to me if you have questions about how any GS features worked or how we solved certain problems. It's been a while but I will do my best to remember.
I'd be more than happy to discuss with you about how Grooveshark tackled specific issues. I'm especially interested in metadata/user uploads management: how did you process user uploaded files and match them to the corresponding metadata, was there some kind of metadata moderation on your side, that sort of stuff.
"Oh, you're struggling with implementing a feature? Well, the guy who originally built the concept your feature is based off of is chilling in the back of the room. Maybe you should ask him."