I think the deal is, nobody but Google can license an orphan book SCANNED BY GOOGLE. That is, Google put the book "back in print" and as a result gets unique control over that scan.
If I understand Google's position, there's nothing keeping anyone else from scanning orphan books, creating their own digital library, negotiating the same terms as Google did, and proceeding to exercise unique licensing control over THEIR scanned image of the same books Google has.
The other side replies: and who would do that, now that Google's done it once?
My take: orphan works are a real problem. It was a great misfortune when the U.S. abandoned its principled stand on copyright (that everything had to be done in such a way that the public domain was clearly defined and constantly increased) and caved to the Robber Barons of Europe.
And Google has tried to address that. Well, it's a good thing for the world, AND a good thing for the copyright holders (if they ever show up). After all, there are no royalties from out-of-print works!)
If what Google says is really true (anyone with a scanner and a friendly library can join the game) and what the opposition says is really true (that there's gold in them thar scans), it seems to me that there's a real motivation for other people to search out yet other libraries that haven't been scanned, to build and license their own competing digital archives.
If there's no gold there after all, it doesn't matter what Google says. Call the library a nonprofit (because nonprofitable) public charity--everyone has to have one for a public image at least, and this can be Google's.
If there's gold, then there's room for competition. Google's efficiency of operations is going to put real pressure on the competition -- remember how long Microsoft's much-vaunted competitive effort lasted! And it begins to matter whether other people can negotiate similar terms.
If they cannot, though, whose fault is it? Google's, for not doing anything one way or another? The libraries and copyrightsholders, for pricing Google competition out of the market through stupidly stubborn bad-faith negotiation?
And is that failure necessarily permanent? Or could another generation of less greedy rightsholders and more efficient collection-scanning organizations introduce competition even if the second attempt fails as quickly as Microsoft did?