I think what Google is doing is fantastic, despite these problems. Some of the sloppiness is a byproduct of the attempts to automate the book scanning and metadata gathering. This has got to be a hugely expensive undertaking, so I can understand the shortcuts to reduce cost.

Until this gets cleaned up, what then, is the fallout from the metadata errors? Researchers will have to verify publishing dates and use other means of search for period pieces, but they will have access to historical texts over the internet. That's still pretty awesome.