Played around with this a little, this is not remotely a proper setup but I have two hs24 sealed at opposite ends of the room so I gave it a go. Very interesting. The room is 17'5" long, rectangular, couple of open doors, one to a large closet and one to the hall. Subs are front and rear center on the floor, distance between subs is 14'. Subs are not recessed in the wall, just sitting on the floor.
-The math is very simple, speed of sound is 1125 f/s, so the delay needed is 17.416 feet / 1125 feet/second * 1000 = 15ms
-Inverted the rear sub, didn't realize to dial down the rear level a little so they are at same volume.
Immediately the very low end is both less strong but also less muddy. Also the response through the room is notably more even. The rear sub just slurps up the bass that would be reflected. This effect would have to be way more effective with four subs front and four rear. It would be interesting to try ported cabs, or even better imagine stacks of horns front and rear
Overall spl, very interesting in the areas circled where you can see the modes neatly reversed and much better behaved. I took measurements in different areas but in general a cool thing is that the dba almost never makes the response worse. You will lose some low bass (and boominess) but this graph is pretty amazing to me, it would probably be impossible to boost that 108hz area enough to have that effect of reversing the null. Guessing that 86 hz is the floor to ceiling mode since the subs are on the floor, a real dba might fix that too.
Couple of waterfalls, the very low bass is the interesting part, and the very noticeable part in room
Here is the before, of course no eq, no audyssey, etc, 3ms delay, in phase
And after, 15ms delay, rear sub inverted