Yes there's definitely the prime vs non prime vs 3rd party price differences.
But are you sure there's no evidence of what he references? He seems pretty explicit, and it seems a very odd conclusion w/no specific evidence: "prices on offer to different customers varying for the same product, presumably on the basis of the customer's willingness to pay more for goods in prior transactions."
When I read an article linked from Hacker News, then read some comments that are generally not tearing it to pieces, I assume the author must have some credibility and therefore is not throwing out conclusions for fun. Also, I have a minor in inferring existence of fascinating undetected evidence.