[looneybomber]

Here's how it works,
+3db efficiency due to doubling the drivers (2x cone, motors, and enclosure volume)
+3db sensitivity due to cutting resistance in half, meaning you get 2x more power (so 500w to 1000w)
Together, because you have 2x drivers and 2x power, you get +6db.

Co-located gives you the full SPL benefit of the 2nd driver...but it doesn't do anything for smoothing out the in-room frequency response.

Having them scattered around usually does not give you the full 6db gain because Sub B is helping to boost nulls created from Sub A and sub B isn't increasing peaks created by Sub A.

Think of it as having two engineers in the same room; You can get twice the designing done (assuming they stop arguing over who's right). Then put an engineer and an artist in the same room; You only get the designing power of the one engineer, but the artist helps smooth things and make things pretty. Wow, that's actually not a very good analogy...

 

[looneybomber]

So +3dB comes from doubling the amount of electrical power being delivered and then the other +3dB comes from halving the acoustic space for each driver.

It's not from reducing the acoustical space of the driver, because they both have the same acoustic space. As taken from http://fulcrum-acoustic.com/wordpress/wp-content/uploads/2008/07/comments-on-half-space.pdf

"The presence of...an identical loudspeaker results in a change in the mass loading (the effective mass of the air in front of a diaphragm), as well as the acoustical resistance (the resistance to diaphragm motion presented by the air). An increase in efficiency implies an increase in the acoustical resistance."

So it should be a doubling of cone area while maintaning an equally proportional BL.

 

[looneybomber]

Well a wall would have quite a drastic change on it's sound; one woofer would be in one room and the other would be in another room. If they were in the same room, that "wall" would not be a wall, but instead could be considered part of the baffle and a baffle change will not have much effect on a subwoofer.

What you're proposing is the same as having opposed firing drivers, one on opposite sides of an enclosure. That's really no different than the drivers together on the same baffle because of how long wave lengths are.