This!!

Not only should audio not be behind a paywall, but there needs to be tracking for listens that corresponds to reads...

Additionally, as the author I would really appreciate being able to select the voice / accent for the reading to match the intended voice of the text / be as similar as possible to my own.