The dilemma has several aspects. One is the gain from making so much written human effort readily available on the internet, both conveniently and hopefully, at modest cost. Against this is the cost that consumers will have to pay Google for access, the amount of which is still unclear. And the fact that Google will end up with a monopoly at selling access to these books, with the minuscule exception provided by borrowing the copy from a library which provided Google the chance to scan it in the first place.
That monopoly will exist by virtue of the government agreed terms now proposed by Google. No other companies willing to scan and offer public access will be in a position to compete unless they can undercut Google's resale prices and Google has the advantage of first-to-market. Libraries for the most part seem to have got little for providing access--those that provided the books for scanning and their members get free access to the Google files. In the end these institutions were funded by the public through taxes (our money) in the first place, so this seems like a clever move to capture the return from a public investment.
But the deal is still subject to government approval. Will the government try to renegotiate the deal? Hopefully yes. The cost, of course, is the delay is making access publicly available. To me, that is a small price to pay.
As an opponent of copyright, however, I would like to see the whole deal and copyright itself abolished. That is unlikely. How about a deal in which unregistered copyrights beyond five or ten years old, automatically expire. And an expiration date on all other copyrights of ten or twenty years. Not very realistic politically, but unless we find some basis for compromise, I suspect we will get no real welfare-increasing improvements.