FWIW, the particular pair of JansZen Valentina that were being used in the latest AXPONA demo had the SE option, which provides omni dispersion at the press of a button on the remote. SE stands for Stereo Everywhere, and when in SE mode, they sound about the same all over the room and also provide some imaging everywhere.
For BACCH, however, they were not and must not be run not in their SE mode, but in their directive mode -- no added ambient sound. The less wall, floor, and ceiling splash, the better for BACCH. This is because BACCH processing is based on crosstalk cancellation (XTC), namely subtracting much of the sound that would normally reach the left ear from the right speaker, and the right ear from the left speaker. Sound splash adds random crosstalk that is difficult to cancel, so it should be minimized for BACCH.
That said, even in directive mode, the JansZens have ±10° horizontal dispersion at 15kHz, so the dispersion is controlled, not beamy. This has the advantage of limiting how much the BACCH system has to compensate for response differences while head-tracking.
To clarify how BACCH's head tracking works, it uses a webcam aimed at your seating position, or if you want your lights off, an infrared camera. It zeros in on facial features to discover where your head is. Calibrations are run for center, leftmost, and rightmost head positions, after which the entire calibration apparatus can be set aside. After that, BACCH adjusts its processing to work optimally wherever you may move within that range.