Really?
It seems like the 1/2" of peel-n-seal is intended as an insulator, and that the insulator works both ways.
In other words: the amount of bracing on a "knock test" of the outer (MDF) cabinet would not be significantly changed by the inner (concrete) panels because of the isolation caused by the flexible layer in-between.
As I understood his design, this was pretty deliberately so as it means that what vibration does occur in the inner panels is not transferred to the outside because the vaso-something layer (peel-n-seal) converts the vibration to heat.
If my understanding is correct (the inner layer is an absorber mechanically isolated from the outer layer) than these gaps give the soundwaves in the cabinet a direct route to the outer shell (hence my question).
There seems to be a misunderstanding about this design, specifically how the peel-n-seal works. It acts as a viscoelastic constrained layer, not as a decoupling layer. By adhering two relatively high mass materials between a third material this design (viscoelastic constrained layer) is able to efficiently turn energy into heat - through high mass and the created shearing force between the two materials and the peel-n-seal, which is an asphalt polymer that is fairly efficient at converting energy to heat and then dissipating this heat.
In contrast, a decoupling layer (what you are thinking of) would effectively remove interaction between the outer shell and inner shell of a speaker. To do this effectively, all physical transfer of energy from outer to inner shell must be eliminated (a pain to achieve properly and clearly not the case in this design). Such a design can be effective, but as far as I know, the inner shell would still require heavy bracing. Such bracing may not be necessary on the inner shell, but I am unaware of any studies that test the transfer of resonance to a driver from the cabinet which may be a factor in such a case.
Quick summary: The peel-n-seal design is specifically created to effectively attenuate resonances while a decoupling design renders them inaudibly by eliminating the chance of transfer to a cabinet sharing space with the listener.
edit: For a simple idea of what a decoupled system would look like think of a box floating in another box - not physically coupled in any way. The drivers would be mounted to the inner box, but form an airtight seal to the outer box, again without any way to physically transfer energy.
