It really frustrates me that fantastic projects like this end up only being made available via some "online catalogue" with tiled zooming and no option to download.
Just stick it on the internet archive, and then voila, the data is actually open, everything gets automatically OCRed, and then we can do fun transformative things like the Internet Archive Book Images project..
All 7ishTB of images are also available on the internet archive, Flickr was a collaboration that allowed browsing them in a vaguely user friendly way. I guess serving over 5 million images isn't trivial and it made sense as a way to help with discovery.