When libraries were started, they were private, then some member-only libraries were born. Finally, the idea of a free and open library for the public was born and the most important libraries follow this policy (Bib. Nationale, British Library, Library of Congress) or something close (university libraries that generally allow public access, though not checkout or stack access).
If Microsoft wants to go and scan the books of the world that are out of copyright and then let that content be available on the BL site, that's fantastic. The Bibliothèque Nationale has thousands of public domain books online already, but they're in image scans, so not searchable. This will be a great improvement. If Google wants to do the same, that's great too.
What I object to in the Google thing are:
1. The hyped up press-release about a beta test with public domain books, and yet all the obvious searches yield sales pitches for unrelated or loosely related books for sale. In other words, they have made an announcement and failed to deliver. It should at this point not be public. We'll see if MS is better or worse when the time comes.
2. The idea that I'm supposed to like Google because they're a "do no evil" company and should patronize them just like I should patronize Ben and Jerry's, whether I like their ice cream or not, because they donate profits to charity. I'll do good in my own fashion and patronize businesses based on the quality of the offerings. Microsoft has no illusions about how they are loved around they world, so they don't act silly about that sort of thing. They expect to sell their products through either quality or strongarm tactics. No illusions. If the Microsoft book offerings are superior, that's where I'll do my book searches. If not, I won't. Simple.
3. Copyright issues. Some books are read (e.g. novels), and some books are consulted (e.g. dictionaries). My books are of "consult" sort, so if a search engine shows a paragraph or two, that really obviates the need for most people to buy the book. Fair Use says that researches can use small extracts in their publications, but Google would essentially be monetizing my content and probably impacting sales. I believe that is in violation of copyright. Why would I publish a dictionary if Google is going to go scan it in and show snippets of 3-4 lines? Why would I buy a dictionary if Google had that available? Microsoft is choosing books "from the older end of the library's vast collection of 13 million titles". These are the books that are most likely to be rare and very hard to find and, of course, there are no copyright restrictions. I see this as providing great value.
Microsoft has put up with accusations of evil and malfeasance for so long, that I actually thing they are
1. just plain smarter on these issues
2. way better at thinking about how everything they do, good or bad, can be spun to make them look bad.
Google needs to learn those lessons.