noble100 wrote: " Duke somehow compensates for this so that the in-room bass response is flatter and more accurate but I'm not certain of exactly how he does this."
According to Martin Colloms and several others, "typical" room gain is about 3 dB per octave from 100 Hz on down. Imo the implication is that a subwoofer's response should gently roll off by 3 dB per octave across the same region. This is too shallow to achieve with an unequalized sealed box even if the Qtc is very low. However it can be approximated with a vented box with the right woofer parameters, box size, and tuning. The Swarm units come very close to this target curve from 80 Hz to about 20 Hz, and since I don't normally recommend lowpass-filtering them any higher than 80 Hz, that seems to work fairly well.
There is another related effect going on with a distributed multi-sub system: The subs are spread far enough apart that their outputs are combining in semi-random phase at the top of the bass region, but down at the bottom of the bass region the room size may be too small relative to the wavelengths for that to be the case, and their outputs end up combining approximately in-phase. The transition from semi-random phase combining of the outputs to approximately in-phase combining of the outputs has the effect of once again boosting the bottom end relative to the top end of the bass region by maybe 2 or 3 dB. If using only one amp, reversing the polarity on one of the subs usually restores balance to the force. If using two amps with the phase controls set for phase quadrature, that phase setting addresses the issue.
Other ways of dealing with either too much or too little "room gain" include plugging and un-plugging ports, engaging the "bass boost" switch on the back of the amp, and using the parametric EQ.
The goal is to end up with smooth bass, and all of these things are just a means to that end.
Duke