[dbspin]

As other commentators have pointed out - the fact that we can do something technically, at the highest levels of competence - does not mean we usually do, or can do such a thing at scale. If you look through the enormous majority of publicly available scans, from Google Books to Project Gutenberg, you'll find low resolution scans with alignment issues, colour issues even missing pages. You're also accessing books at the whim of the legal and technical infrastructure you don't control.

Sure you could keep a secret digital archive to pass hand to hand for circumstances when (as has just happened to the internet archive) thousands or millions of books are suddenly banned. But you're relying on access to open computer architectures that don't phone home your reading of those books (like Microsoft's 'Recall'). You're reliant on power. You're reliant on compatibility into the future. With a printed book, you have a physical object, duplicated hundreds of thousands of times and distributed. A physical book can be read by daylight in a refugee camp, a war zone, a desert.

It may seem a silly example, but if the Roman or better yet the Aztec empire had written their manuscripts digitally - we likely wouldn't have a single record from that era. They would have been erased through chaos or deliberate eradication. Don't be so quick to dismiss the physical.