When news came down the pipe that Virginia was going to have a "Standards of Learning" type exam for colleges, the first thing I did was jump on board to help scope (and my plan was to help author) the test (I was physics faculty). I'm no fool: if we're going to give the administration a vehicle this powerful, I'm driving.
On the other hand, when the Standards of Learning for K-12 schools were first implemented, a remarkable thing occurred in education: teachers started talking about teaching again. Until then, every conversation with them was about how bad the administration was, how difficult parents were, how unprepared and unmotivated the kids were, etc. When those tests went in the conversation turned to: how in the world am I going to teach all of this?
How in the world, indeed.
You are also correct that the way peer review is implemented leads either to back scratching or very poor morale. The way to do it is to get folks from one institution to review folks (as a team) from another. That would place a huge demand on the personnel, though, and I think you would have trouble getting it in place.
The problem with that arguement is that teachers are hostile to every form of evaluation that has been proposed and are unwilling to propose anything that they'd find acceptable.
> Peer review is the ideal.
No, it's not. That's like asking ford to review chevy. It leads to backscratching.
> it certainly will create cliques of teachers, but only rarely cliques of bad teachers.
You clearly don't have any experience with teachers.
> those facts that are deemed convenient and useful by powerful cliques in the state legislatures.
Those facts have the interesting property that they're at least nominally under the control of the folks paying the bills.