What I find interesting is trying to prove harm done by the fake reviews. If they're against the terms of service, okay, ban the accounts, and take some back-room steps to monitor if those persons try to rejoin (e.g. limited utility of IP address focus). I don't think we're talking about criminal masterminds here.
I'm kind of joking here, but I'd almost try out the defense that paid five-star reviews are a "necessary evil" to counter the inevitable "idiot with a keyboard" which enables any wrong-thinking person to write a one-star review. What I mean is will this case include evidence that non-paid reviewers are just as damaging - if not more so - than paid fake ones. Or, in other words, can Amazon prove genuine harm via "fraud" is able to be differentiated from "stupid people saying stupid, wrong things" in the comment environment.
Personally, I think it's an "in-house" Amazon issue where they should be 'moderating' comments, but that takes labor and effort; if I've learned anything about Amazon's view of menial labor tasks such as fulfillment, I don't think anybody would want the job. If there's better money in fraud than policing the fraud, then maybe the business model should be re-evaluated. I'm not trying to justify what is done for hire - to me it's about on par with "essay writers for hire" and I hate that shit - but it does make for an interesting problem of commerce.