I wonder about the meaningfulness of reporting by students based on memory as data for the percentage of teachers who are "great" and "awful." Teachers who are really good are often very demanding of their students, and students frequently don't cotton to being challenged -- realizing only years later what it did for them. And we would need to know much more about the student who thought 40% of his teachers had been awful; one wonders what sort of student he was, and whether his expectations he had of his teachers were realistic or appropriate.
Ah, the old evaluation-resentment-payback nexus. I've never trusted the results of student evaluations, unless I knew in advance the qualities of the students themselves.
To be fair, in high school the teacher most disliked, at least in my experience, was the one the students could take advantage of. No one ever took a teacher seriously who could be peer-pressured in my class.
And, of course, students are free to dislike good or great teachers, if that's their thing. But the conflation of liking/disliking/evaluating is fatal (and, I think, endemic) to the evaluation process.