I worked in the music industry (in IT). I have no idea where the idea came from that the music publishers didn't have to renegotiate contracts to get digital rights to the music. In reality, when digital rights became important, the music companies spent a huge amount of time and money having teams for at least a decade tracking down rights-holders and negotiating digital rights in order to sell their back catalog, and of course made sure that their new contracts covered selling through the digital service providers. Book publishers have essentially the same legal challenge (though admittedly the details are different).
What is really different is the production logistics.
Music has been digitally produced for a very long time, using open standard formats, and for pre-digital material it's relatively easy to digitize audio (and video) from master tapes, so you only need to do "work" to deal with some very old, obscure media, which is only done selectively. And the music publishers have built systems that are very, very good at managing and format converting huge libraries of audio and video. So, 99% of the time, digitally selling back-catalog music and video is logistically fairly easy - QA, package, price, and send the files to the digital service providers.
Books, however, have been authored in a series of random formats, and for older books there's only the physical book or manuscript and nothing digital. Which means that you often need to physically scan every page in the book/manuscript, OCR it, clean it up, QA the result, etc. And even for the digitally authored books, you need to track down whatever specific physical media and formats each publisher or author used (MacAuthor on 3.5" floppy, LaTeX, MS Word 3 on 5.25" floppy, etc.). So, overall, physically and logistically really complex to deal with for every single back-catalog book.
Look at what Project Gutenberg has produced - an amazing collection, but it required a massive investment of (volunteer) effort to process the books into digital formats.