Gabriel Power emails me:
I have not seen anyone discuss this possibility of a hybrid open/classic referee process: Editors approve set of X reviewers (say 300). Then, every submitted paper can be reviewed by any approved reviewer. Editor then considers all reviews. If no review, then editor assigns or desk rejects. Issues: Incentives? Selection bias? Matching author quality with reviewer quality? Conflicts of interest? Pros and cons? I think it is intriguing. Sincerely yours — Gabriel Power
I would stress the goal is not to find the “best” reviewing system. Rather we are looking to set different reviewing systems in competition with each other.