Speaker cabinets: How important is inert

"but by itself, a cabinet that is simply dense doesn't address the issue"

Actually it does address the issue. Not 100% But generally. A speaker has a specification called vas, this is the Equivelant air volume or the volume of air that has the same compliance as the speakers suspension.
The suspension i.e. voice coil assembly, spider and even cone material affects how the driver reacts to the air behind that driver. That is why In a sealed box a drivers frequency curve will develope a peak as you shrink the box size or develope a dip as you increase the box size. This happens on any traditional cone type speaker regardless of size. Cone breakup because of that back wave is what causes the smearing effect that you speak of. The right cone material will not have that smearing effect of back wave. In general, cabinet stuffing has three purposes, to break up standing wave, to cut resonance and to trick a driver that it is in a larger box, by changing the time that it takes air to travel within the speaker box.

Given there are hundreds and hundreds of speakers on the market, there are obviously a variety of opinions as to what is the best approach to speaker design.

The Vas parameter is used in determining the bass performance of the speaker. One can model a number of different box volumes and see a significant change in performance under 100 Hz but virtually no impact on the performance above. If you reread what I wrote above, it is this range above the bass where it is often difficult to control the back wave re-radiating through the driver cone material.

Of course, cone material can be selected that is better in this one area, but that cone material may represent a loss of performance in other areas. Almost all engineering efforts represent finding those compromises that result in the best balance in the opinion of the designer. And, as is routinely demonstrated on this forum, there a considerable range of variety in those opinions. ;-)

"One can model a number of different box volumes and see a significant change in performance under 100 Hz but virtually no impact on the performance above"

Hi Mlsstl,
I was only addressing cone speakers. VAS affects all cone speakers in a box regardless of size. It is just as important to have your midrange in the correct box size as the bass driver. Its final Qts figure will relate to a hump, flat or a dip in its frequency. If you develope a hump, it would easily portray a coloration or "smear" in frequency. As far as controlling back wave in the cone material, I was specifically speaking of mid range drivers. Back in the old days of paper cones, we treated them with Latex to stop cone break up, Later came poly, kevlar, carbon fiber, aluminum,....etc etc etc.
The latex helped control breakup. The more dense materials normally were not as effected by backwave coloration because of the stiff cone material. What I am saying is the cone material and box material are not intertwined. They both matter greatly, but independently of each other. Sorry, no disrespect intended.
Back to Enzo618. Any material that is rigid and dead would be great. Normally thinner materials require a bit more bracing, but can work. You've probably seen concrete enclosures. I've not experimented with alot of materials, I've used some fiberglass, but it had to be thick and re-damped, I like your epoxy sandwich idea.... but in the long run, a proberly damped inert material that has the proper air space will do a reasonable job. Some may be slightly better than another, but the standard braced MDF is hard to beat.... Dollar for dollar, I'm not sure that you can.

No disrespect taken (but then I've never worried too much about that issue one way or the other...)

I'm with you on that brother