IME negative feedback can be the most valuable kind, because it points at where you can improve, but you have to also take it with a grain of salt because some people are difficult to please: this feedback has to be managed carefully - e.g. by allowing those involved to have their reactions to that feedback included in what management see.
Yes this is definitely interesting, but as I have currently zero experience in actual repair events I don’t see how I can directly get involved. Measuring long-term impact/behaviour changes must be much more difficult than getting an immediate response to a specific repair event. To make the immediate data collection as lightweight as possible, while you could use one of those slightly irritating devices you see at exhibitions with three big coloured buttons for and , it might be better if e.g. the host gives the repairee a physical token (plastic disc) to put in one of three collection boxes, also ensures only the repairees give that aspect of the feedback.
Being able to give some sort of small/relevant reward for continued involvement might help with getting longer-term feedback but I’m sure a statistician should be all over this to make sure the bias in self-selection is taken into consideration. I imagine it’s rather like trying to change the vector of a ship by rowing to push it sideways: for the effort put in by any one person/event, the result is miniscule/unmeasurable but combined over geographies and time there should be a slow trend.