What I dislike about this method is that it seems to be focussed on A=440 Hz, which is arbitrary. I assume that if the learner drifts later in life by under a semitone, then things will seem like they're between keys.