The only thing you know for sure is that the polar response will be symmetrical. But even though it's symmetrical, it's going to be very complicated. Much opportunity for comb filtering and various lobing patterns to confuse the power response.
The analogy to a conventional MTM speaker is not bad. You see the same issues with MTM speakers, but just not as involved. In my opinion, the only valid reason to construct an MTM speaker is the sensitivity issue since two conventional woofers in parallel usually match fairly closely to a single tweeter sensitivity-wise and it yields an approximate 6db SPL increase. However, a standard MT configuration is preferable for all other considerations. I suspect that might be the case with this experiment as well.
Cheers,
Dave.