As a manager I can honestly say that the peer comments and ranking give me better insight into things I might have missed. I don't follow people who report to me around all day and see every little thing they do. I get a sense of things they do that I didn't necessarily have visibility into.
This can totally move the needle for me. It's not going to change my opinion 360 degrees but it can move the needle. In general I'm going to put a decent weight on peer input.
That said in any system when there are enough individuals, managers or non-managers, that actively try to thwart and game the system, things are not going to end well. That seems to be the kind of place Rachel is describing. In that case nothing is going to work.
To Rachel's example of moving Mount Fuji with a teaspoon. Is that something you were supposed to do? One of the other comments says that if not that reflects poorly on the manager but when you manage senior people you don't manage them at the level of one teaspoon at a time. A senior who chooses to move Mount Fuji, and is given the right input, can totally be penalized for doing the wrong thing if that was wrong.
EDIT: All that said I think I read a paper somewhere about how performance reviews are generally counter-productive. I've always them somewhat disruptive as an IC and not a lot of fun as a manager. That said, there has to be some method by which we determine how much responsibility and how much compensation (e.g.) different people get.