Assuming the kind of books listed here are the kind on the daily deal (1.99): The average user would have to read only 4-5 books a month to break even (say 9 books every two months counting taxes).
I doubt most users would read that consistently. Maybe the first month, but usage would likely drop, and the deal works out better in Amazon's favor until they quit. Which for many users could be until their credit card expires.
I probably only read 6 books a month at the most these days, and some months I probably drop down as low as 1 book every month several times a year. I'd probably enroll in this service so even if they have 9.99 normal cost ebooks, I think overall it'd work so Amazon doesn't lose much more than they would with their current structure.