We don’t normally model weird-o requests like this. :D
The question is really whether the internal air suspension would continue to work as a simple spring, and the answer is maybe. Depends.
However, part of your question is easy to answer. The sensitivity of the driver has nothing to do with the enclosure. It’s all in the magnet, voice coil and suspension losses. This is why (contrary to myth) bass reflex enclosures and sealed drivers have the same sensitivity. What’s different, when either can be used, is the -3 dB point will be lower, given optimal enclosures for either.
Put another way, the enclosure size (bass reflex or sealed) only affects the low end of a driver's output. The mid to upper end of the driver's output "ignores" the air space behind it. At these frequencies the baffle dimensions and type however do come into play.
Having said this, weird-o internal baffles may have frequency dependent effects that may be best built than discussed. I’m sure there are ways to create frequency dependent effects due to reflections and constructive and destructive interference. It is quite normal to have internal bracing with lots of room for air flow to create stiff enclosures but we must remove their volume when calculating the optimal enclosure space.