It's hard to compare to books, look at the sales numbers and point to the cover as the difference.
I can imagine methods to test the impact of a cover. Take two groups of people, and show each group a list of books. All the books in the list would be identical, except for one: one book has a different cover. From there you can do different things. One thing you can try is to offer the subjects one free book, and see if the cover made any difference in people's book choices. Another would allow them to have one free book, but they could only click on a couple of them. This would test the effectiveness of the cover in getting people to take a closer look at the book.