Sorry if this seems going further off track, it is obviously not my intention.
Nice find indeed. As to the validity of the data, you may notice that all the questions are pretty much behaviorally anchored, thus less likely to be biased by personal liking of the coaches.
Sampling bias certainly...