The input signals would have to be not 120 degrees apart (like 3 phase power) to get an output more like 22 degrees apart.
My thinking is as 1 coil runs out of magnetic flux in the gap the other one still has plenty to push/pull the speaker cone movement at either end.
On the crossover point of a cheap class b amp. I'm not sure if this 2 phase setup makes any difference because the distortion maybe halve but it is there twice in the cycle. What difference it makes I cant tell yet with my setup.
I have only a triple 4 bit D to A converter at the moment so I am only looking at bass frequency's and one of my amp refuses to drive the simulated load of a single coil speaker (sub) wired with resistors so I have done 2 phase tests so far.
Here is a pic of the Audio side also there is 3 computers in the setup as well. Had to wire 3 PC power supplies up as well to get more current.
Sent a few emails to speaker repairer's and it seems a 3 phase voice coil will cost a real lot so Im looking a 2 phase at the moment eg a speaker with 4 layer coil.