Great comments, Gianluca. What if we are making measurements relative to the particular player, so that what we are making measurements that help identify differences and preferences relative to that player? Aside from that qualification, won’t the measurements aid in grouping different bands or models as being similar in sound?
If I say that among five melodicas, the Juliette Pianola is the one that sounds most like the Borel Clavietta, wouldn’t oscilloscope readings help in trying to confirm or disconfirm my claim? If so, would it do so only in relation to the person who is playing during the test?