I think Blount is overestimating the capacities of text-to-speech technology. Sure, there may be speech synthesizers that can cough and say "umm" and do other human-cadence-like things, but I suspect that inferring all those details of intonation from the normal punctuated text of a novel is beyond the capacity of present technology, and may even be AI-complete.

If the digital files on the Kindle contained lots of metadata, added by a human editor, to assist the text-to-speech system, then Blount would have more of a case.