I think this is a neat idea. Basically, you'd get more "credit" if you're cited by a good paper than if you're cited by a bad paper.