I'm sitting here listening to my 91" tall horn stacks and pondering the sound. I think the perception of dynamics even at low listening levels is largely caused by room interaction issues due to the directionality of the speaker. Sensitivity or efficiency may just be a surrogate for dispersion characteristics. Anybody know of a low efficiency speaker that's highly directional? Electrostatics! Except they have a back wave, but if they are far enough from the back wall they should still have a dynamic jump effect before reflections come in and reduce the buildup slope. This is the two streams concept. There needs to be enough time delay between the initial sound that reaches the listener from the speaker compared to the onset of reflections that cause sound level to continue to rise, but more slowly. Buildup of sound in the room always sounds "slow." Reverb never sounds dynamic, and if it comes on too soon it muddles up the perceived dynamics of the speaker.
That being said, I don't get the perception of superb dynamics from electrostatics or Magnepans I've heard - except some little desktop Maggies at close range. Those were pretty intense. So maybe a point source works differently than a line or plane source in how it loads up a room. Point sources have better fall off rates, so that's one possibility. My bass horns are stacked up as line sources and I have to admit they don't seem able to produce the punchy, slap you I the face and make you blink kind of bass I have heard from much smaller speakers I once had in a 6.0 surround configuration in the same room. Sitting relatively close to 6 little woofers that were surrounding me produced astoundingly tight and fast bass in that particular situation.
To evaluate the perceived dynamic speed of a system I think we need to consider how the speaker is interacting with the room and how much time we have between the sound immediately rising at the listening and the onset of the slower room buildup. That's my proposition. Immediate thoughts on how to test this - make a super wide dispersion horn system. Maybe add some sideways and backwards facing horns so it's an omni horn. If it sounds just as dynamic at low volume as the directional horn then my theory is weak. Another idea is to make a super low efficiency highly directional horn system. Mass load a compression driver and stuff the horn with foam, and add a load resister to make the amp work harder.
Right now I'm running a JBL2426h on a BL-409 plastic waveguide all the way down to 600Hz with a 1uF cap in series. That puts the 600Hz point 30dB or more down in sensitivity. There's also negative EQ added to extend the top end response. The driver is still super efficient because it's not drawing much current from the amp, but it is definitely not sensitive, as can be seen by the fact that I have the tweeters turned up 11 dB at the pre-amp compared to the woofers, and the woofers also have negative EQ added to them to extend their low frequency response. This doesn't rule out efficiency as a cause of dynamic sound but it does rule out sensitivity because if anything they sound more dynamic to me with their improved directional control and smoother response.