Let me ask you this: would you rather buy one of the first 100 prints of Djikstra's "A Discipline of Programming" for $1000 or would you rather buy 200 Nancy Drew novels at $5/book? Books are not, to some extent, able to be substituted. Yes, if you are looking for a good vampire erotic fiction novel and are choosing between two unknown authors and one is priced at $15 and the other at $10, sure you would get the one priced at $10. But if both are priced at $10?
The point is that books are not bought as a simple commodity. Nobody says "I am low on books, let me go grab a random handful, based solely on price." No, people buy books with good reviews and by recommendation. When you are viewing volume book sales from Amazon's point of view, sure they all look the same. However, taking a closer look, a publisher might be able to price books more efficiently. If Amazon has helpful data to add to this, why don't they open it up and let everyone, publishers included, see it to help price the books better? Presenting a single data point in the huge possibility space of book prices is not a valid argument. Even if you let go of all the economics involved and stop thinking about how people buy books, don't you think that that ratio of 1.74 would be different for books priced at $12 and $17? Does this vary by month? Do certain genres have different price elasticity? Amazon is not trying to be helpful here. They are trying to squash publishers, gain control of the market and dictate prices, which they cannot currently do.