I'm sure some of the builders will chime in on this, but I'll throw my theory/experience into the pot. Bearing edges, head construction, tuning & venting are the elements at play here. With the exception of extremely thin vs. extremely thick shells, I don't think the shell material makes much difference, except to dictate the fundamental tone.
Think of the head as a membrane tensioned over a ridge. A sharp bearing edge offers less impedance to free resonation than a more rounded profile. Let's exaggerate the rounded edge, & imagine it as a flat edge in profile. When the head is struck, it moves up & down, with greater throw in the center, then diminishing towards the outer edge. If the edge was flat, the head would only contact the outer edge of the flat on the rebound stroke, but contact the inner edge on the down stroke. This change in contact position, or fulcrum if you prefer, produces a dampening affect. Ok, that's an exaggeration, but relates to the difference between a very sharp, & more rounded bearing edge profile. Effectively, I'm encouraging you to imagine a microscope image of the dynamic about the bearing edge. Put simply, the greater area of contact surface between the head & bearing edge, the greater the dampening affect. This change of fulcrum also explains why Caddy's drum chokes when struck hard. He's essentially increasing the amplitude of the oscillation, thus bringing different fulcrum points into play. A head is at it's most resonant when struck very lightly and resting on a knife sharp edge. Sorry, this is a bitch to explain in writing.
Then there's tuning. Two heads of the same material, & tuned exactly the same, with identical bearing edge profile, will offer the greatest resonance. Any differentiation from that will produce a muting affect. Sympathetic resonance is at it's most pronounced when two identical elements react reciprocally. It's not all about creating pressure waves in a sealed environment either, hence the influence of the breathing hole. Relying totally on a captive pressure transfer confuses the natural sympathetic resonance of the two heads. Put simply, the returning pressure waves collide with the mating head, but not necessarily in sync, hence the muting affect. Remove some of that pressure influence, & the unimpeded sypathetic resonance is allowed to dominate. A clear sonic image is transferred perfectly without the need for containment. There's plenty of examples of this in real life. Think of how your ears react to speaker output.
A similar & related explanation also works for two ply vs. one ply heads. As already pointed out, both plies are not equally tensioned, but more than that, it's a composite form that reacts less harmoniously than a single element.
My next drum project relies very much on the principal of sypathetic resonace being more useful & dominant than resonance transferred by point contact. I'll be swimming against the tide on that one, but the results will dictate who's right. I find the prospect of being completely wrong adds a lovely element of anticipation.