chuckle recall, a paper is given to 3 or 4 reviewers. No need for hierarchies of reviewers; it’s more like a jury; if all the reviewers more or less come to the same conclusion we can have a high confidence that the decision is the correct one.
Under the proposed plan, if one of the reviewers gave a review which was radically different, or otherwise obviously slap-dash job, payment could be withheld and another reviewer commissioned.
many CS conferences have something literally called a "meta-review" and then there are further senior people who read and oversee the meta-reviews. it stops there though.
Or possibly just a way to review the reviewers. This opens itself up to competitor bias, though, so it would need to be thought out in a way to minimize that.
