I don’t see the problem. We should encourage publishers to move from real DRM to this kind of scheme. Is it an easily defeated system? of course it is. It doesn’t have to be watertight. All it has to be is something that will keep piracy in check by making non-techies less inclined to upload their purchased books to Pirate Bay.
I mean, look: using this scheme you get a book you know you can read anywhere or even sell to a third party (assuming said third party keeps it to himself). It doesn’t hamper readers in any way. And seriously, the scenario in which someone steals your laptop or phone and then goes to upload your books is far-fetched at best. Thieves don’t care about your media collection.
For those who worry that this might introduce errors int the text, well, ebooks are basically a bunch of XHTML files with some metadata and other files. It is trivially easy to instruct the variance-introducing process to change things that are only visible in code view and don’t code for any visible difference in the end product, such as multiple space characters, random line breaks , introducing random tags that don’t code to visible text, etc.
