So, basically, you want automatic (or semi-automatic) population of track data, based on information already in database?

This is interesting and probably not so hard to implement, but my concern that it's hard to verify accuracy of auto-populated data. If there's any error, you probably won't discover it until much later. Also not clear what to do if there's conflicting data in database - e.g. different credits on different albums (typos on covers, etc.) Still, I like the idea.

Andrei