The waves from inside the midrange cabinet will be transmitted outside through the cone of the midrange driver, which is very thin compared to the cabinet.
You sidestepped, you assert several times the midrange cabinet is somehow lightly stuffed. It's stuffed to 100% of it's volume. Are you suggesting packing in material to neutron star density is going tame some remaining imaginary issues?
Yes you have to be sure that the material doesn't get contact with the cone. Such polyester wool has almost no effect in the lower midrange. You either have to stuff it or use much denser material. It is effective in the higher mids though.
I actually have done this experiment before. Let's do it again, since what you say is inconsistent with prior measurements I have made and seen.
I have a midrange, a cabinet, and some Jute and Polyfill stuffing.
The midrange is easy to remove with threaded inserts. I used a driver, set the torque to 11
which gives a repeatable seal.
I did 5 treatments, the first control is with no stuffing. I did runs with single layer of Jute, and the same with single layer Polyfill. The quantities are typical cabinet treatment, but not filling the entire cabinet like Genelec. I also doubled the amount of Jute. Last is with double Jute and all the Polyfill I could fit. The pictures below are of the actual cabinet stuffing for a few runs.
I measured both the impedance and frequency response. I marked the position of the speaker carefully to get some level repeatability in the FR measurments.
Here is the FR in the cabinet with no stuffing:
The impedance trace shows resonances at 770Hz, 1kHz, 1.9kHz, and 6.9kHz. The 6.9 kHz is the classic cone breakup mode. The others are less clear.
Looking at the distortion of the driver in empty cabinet, the 6.9kHz breakup mode shows ramifications in the distortion, with a peak in 3rd order HD at 6.9kHz / 3 = 2.3kHz, 5th order peak at 6.9kHz / 5 = 1.38kHz, 7th order peak at 6.9kHz / 7 = 986Hz. etc...
Let's see how this changes with damping:
Once you line the cabinet, the resonance frequency response irregularities at 1kHz and 1.9kHz go away. Jute or Polyfill, even stuffed to the gills doesn't dramatically change the response. Double stuffing the cabinet with Jute and Polyfill did reduce the bass by 1dB below resonance, not exactly a good trade for no improvement in midrange performance.
Looking at the distortion with single layer of Jute, you can see the resonances at 1kHz and 1.9kHz clean up in the 2nd HD trace, the odd order HD components are mostly the same though.
This makes sense, the cabinet is 7 inches deep, and 4 inches wide, the Jute is enough to break up the standing wave. Note, this is not radiation through the driver's membrane as you assert. That occurs too, but is not evident in the measurements, i.e. is so small it should be ignored. So your assertion about resonances from broadcast through the driver makes no sense, and the actual resonances due to cabinet reflections are tamed with minimal treatment alone.
None of the treatments make any difference in distortion performance above single layer of Jute or Polyfill:
None of the treatments make a difference in midrange impedance resonances compared to single layer of Jute or Polyfill:
Stuffing amount and type does affect the woofer/box resonance at 118Hz, lowering the impedance peak and slightly lowering the system resonance as predicted (see for example
Vance Dickason's Cookbook, or any box simulator). You can see that the game of lowering the bass resonance by overstuffing is counteracted by lowering of system efficiency after more than single layers are applied, so the often-repeated statements about 'improved bass performance' with large amounts of stuffing is also a fantasy. Somehow that myth needs to be put to rest as well.
So no, Genelec (and other manufacturers) aren't skimping or putting less than optimal amounts of stuffing when they line a cabinet with a single layer of jute or polyfill or whatever. That myth of the understaffed cabinet doesn't apply in this case. In fact, the Genelec is filled completely, and very tidy, which you don't seem to acknowledge. They are avoiding doing something silly that has no practical effect, even drawbacks. More stuffing doesn't improve sound, well designed cabinets improve the sound Minimal fill tame reflections. There is no magic in stuffing cabinets, despite what the unsupported statements around the internet says. And if the cabinet has a mechanical resonance, stuffing isn't going to fix it. And no, the sound leaking through the driver membrane isn't a dramatic effect that is audible. You are tossing ideas, but not measurements.
Lastly, here is a different cabinet I made, with a diffuser on one wall. It tames the mode between front and back of the cabinet, does nothing for the other dimensions' cancelations, and was a pain to build.
Tweaks like this are of no value, and distract from the already difficult task of making good speakers.
