First off, I don't think that a teacher's performance eval should be entirely rooted in state tests, nor do I believe that the current method of testing is by any means perfect, but I think value-added measurement is a decent measure of how teachers are performing and how much knowledge students have gained.
What we're looking at here is an extreme case where that value added measure has no more meaning and the actual evaluation has no statistical meaning (the students performance was already top notch, they were probably more likely to score a little worse.)
What SHOULD have happened is that they should have been evaluated using a more rigorous / advanced curriculum that was aligned to what the teacher was teaching (ie, high school algebra). The system is not perfect, but it's not terrible--they just need to evaluate the system, adjust, and iterate, something many here are plenty familiar with.
(Disclosure: I teach at a charter school)
Right now it sounds like they are awarding the teachers based on the percentile level of their students compared to other students in the same class level, which is pretty much guaranteed to be useless noise if your students all start at the 98th percentile. It would be much more meaningful to be able to say, "Each of your students clearly knows more algebra than he did last year."
The other issue is how to motivate the children to actually want to do well on the test. Somehow rewarding a kid who does better than he did the previous time he took the test seems like a good first step...
The other problem is that this is a problem at which we should be throwing our best data scientists, and in many cases we're lucky if those involved have an intro stats course under their belt. Evaluating student performance is no trivial task (ESPECIALLY when trying to create some accurate measure of reading level), but we're not putting the brainpower behind it to figure out. And as much as I love it, I don't think Khan Academy will be solving those problems, but I'd love to be wrong.