Lets say 3 in 100 reviews are good. And you want those 3 to come back and review the next product. You are even willing to pay them something. Is there a guarantee that they have interest and will show up, and if they do, is there a guarantee that they produce a useful review of the same standard the second time around?
Thats just one simplified reason about the dynamics of this process, that is going to block everyone from getting things for free.
Its easy to react, to push to dismantle it and you just end up with YouTube style mega mass of comments under every vid where the best comments are guaranteed to get buried unless they pander to the will of the herd.
It's more like whether Amazon should pay the people who decide what should be in the Amazon Basics line, which I suspect they do.
But anyone who has tried knows its very hard to do for free. Just ask the mods of HN whether they want to work for free and the costs to them to maintain a standard.
One can imagine a world where high-quality reviewing entities exist and attract top researchers without being profit-driven publishers.
Sometimes people agree to review articles because they get to see research before everyone else.
Journals extort hundreds of man hours per conference from reviewers. The largest cost of all, research and writing the paper, isn't reimbursed in the slightest and largely covered by tax payers, who then still won't be able to access the results because of paywalls.
The entire business plan of journals is to take research and peer review for free, and then sell it. Back when distribution was costly that may have made sense, but nowadays with the internet the cost of distribution is a fraction of what it used to be.
People who think open access is the solution to that are dealing with a focusing illusion - https://www.edge.org/response-detail/11984
Amazon reviews are not remotely a good comparison. There is huge incentive in terms of experience gain, getting a bit of a sneak peek at research, and reputation gain(lets be real, its not that uncommon for people to be fairly confident they know who an anonymous reviewer etc is).
I'm not trying to make any point other than to add further disagreement to the idea that these reviews are not currently obtained essentially for free.
Additionally, published papers are often cited by other papers and peer-reviews are necessary to prevent a false premise from being propagated through citations unlike products on Amazon.
Peer review is not (if you want to keep your career, that is).