I think that's a good illustration of the quandaries of SOPA: the problem of figuring out these nuances of ownership without disrupting the flow of information. It's easy for a publication to determine whether they themselves are complying with fair use. Not so much with Google and it's automated high scale systems