I understand it presents legal complexities for licensing, but watching captions still isn't the same experience as a book.
More, there are people out there for whom visual books don't work and audiobooks don't work but would be able to use a book with both in tandem.
An automated caption system for an audiobook is basically that in reverse.
But please don't throw out the idea of autocaptioning audiobooks as unnecessary or only a vehicle for piracy.