What's scary about this to me is any implementation of ContentID will be based in some sort of score way, determining the likelyhood that something is infringin

What's scary about this to me is any implementation of ContentID will be based in some sort of score way, determining the likelyhood that something is infringing. These scores would be completely arbitrary, but you should be able to imagine the outcry if it became public that Google set the score for removal to 1000, but if they had set it 5000 would have blocked 5 times more copyrighted material. The public/politicians wouldn't understand these numbers are both made up (and so any difference between them is really meaningless), and would lead to more blocking of legal material.