This sounds like a valid viewpoint, but it could lead to significant downsides. Like pay per review scams on Amazon. I don't know how this policy will play out in reality.
Nothing would change (except that Elsevier etc would have to reach their pockets for the first time).
The editor invites domain experts to review manuscripts. By compensating the reviewers it does not mean that you suddenly get more domain experts. Same people will be invited. But they will be compensated by taking a cut of the publisher's profits.
Or we can cut the middle men and publish in community maintained non-profit open-access journals.
