I think you underestimate the ability of students & parents to judge teaching ability separately from likability. The users of a product are best suited to know what works & what doesn't.
At my high school we had a mix of good & bad, as you'd expect. Everyone knew which teachers knew their subject, could communicate well, was more dedicated etc. Some of the most engaging teachers were the least liked because they graded harder, but if you had a rubric that included "grades fairly", "not boring" etc, students would be able to correctly assess them.
Of course some would be dishonest, but noise will cancel itself out, by and large.