While you are right, that there is not decent empirical studies which show vouchers improve educational outcomes, and I am skeptical that they would, I'm slightly dubious as to your theory of slow "feedback" as a reason for why they don't.

While teaching outcomes are hard to assess in short amounts of time, and teaching outputs are also hard to evaluate in the short term, if you believe that organizational culture has any effect on those outputs/outcomes, and thus there are such things as "school effects", within a generation there should be systematic differences in the outcomes of students from different schools, which could then be used to inform better choices on the part of parents.

But you are right, those good things, if they exist, will take decades to observe, and there is no reason to think that vouchers will do much good for current students.