From what I understand, to work with copyrighted books they need to essentially format shift (i.e., scan and destroy the physical book). So a book vault would not solve this issue.

A book vault would still be useful for out-of-copyright works, but this would only cover a (probably relatively small) portion. Also, I'm not sure how easy it is to reliably determine copyright at scale, so they might just decide that it's not worth it.

At this point my only hope is that in the long run these scans make it to the public somehow (leaks, copyright changes/expiration, whatever), where they can then be accessed and preserved by everybody. Then we could have our true digital library of Alexandria.

If they’re old enough and rare then copyright is moot, as you mention. The size of the vault is of interest but it’s not something that makes the vault possible vs impossible.

I’m not a lawyer, but I would guess the theory behind the destruction is to guarantee that they are authorized under fair use to have and to use their digital copy. Imagine a situation in which someone else becomes the owner or is using the original. If you destroy it, that becomes impossible. If that’s part of the constraints, then a vault, rather than a library, could be a key distinction.

You might be confusing the need for an original work that draws upon another to be “transformative,” in order to not violate copyright, with the notion of “fair use,” which allows scanning something you own for your own personal use.

They could put everything in some kind of nonprofit book vault/archive that includes all the source pages as long as they didn't use or distribute it. They could even provide a mechanism for rights holders to recover the text for free, if they need access.