Care to point to some of those studies? How do they define what a "very good teacher" is versus a "mediocre teacher"? Interviews with students? Interviews with teachers? Average results of their students?
The studies in question are stuff like: if you look at standardized test scores before/after each particular teacher's class, and then run statistical regressions throwing in a bunch of other variables, you will find that most of the variance in the improvements in student performance can be attributed to factors other than which teacher they drew.
But this is a weak way to study the most important influences teachers can have which are about inspiring and exciting students and might not be particularly observable until years or even decades later, when e.g. a student with an inspirational middle school teacher chooses their college major or career path, which might be a different choice from the counterfactual student whose teachers were just phoning it in.
But this is a weak way to study the most important influences teachers can have which are about inspiring and exciting students and might not be particularly observable until years or even decades later, when e.g. a student with an inspirational middle school teacher chooses their college major or career path, which might be a different choice from the counterfactual student whose teachers were just phoning it in.