Interesting perspective on it by Dan Cohen. Main conclusion: the devil's in the detail and we don't have those yet. But he highlights the fact that the deal seems to allow humanities scholars computational access to the scanned works. That could be a big deal. plus, of course, the fact that the deal seems to allow any library that pays a subscription terminal access to full text of everything in the database. If true, that changes the game.
Just for interest, here's a piece I wrote about the Google Books project when it was first announced four years ago.
My most recent comment is here.
Importing Functions From DevTesting Jupyter Notebooks - One of the ways I use Jupyter notebooks is as sketchbooks in which some code cells are used to develop useful functions and other are used as “in-passing” ...