I wonder what the effective order is in relation to its sensitivity pattern selections, the quality of the off-axis nulling, how well artifacts are suppressed, how noisy, how natural it sounds..
Interesting choice in that it appears to provide simple analog output only. Sort of like an old-school LD condenser mic that provides variable pattern via a required outboard power supply in that regard - a cool option for simple use, yet unfortunate if that means the pattern cannot be altered after the recording has been made. Maybe USB output connection provides that capability.
The second output (labeled null, equating to everything outside of the primary output pattern) is one of the more interesting things. I've thought about that in regard to ambisonic recordings, but don't know of it being implemented in any ambisonic tools. It sort of falls into my philosophy of recording sound from all directions collectively, via discrete segments designed to combine gracefully, as a useful way of deconstructing the acoustic upon recording with sufficient degrees of freedom that it can be reconstructed in a perceptually convincing way for improved reproduction. Similar to the OMT implementation of a rear facing microphone channel, but differs from ambisonics and my broader approach in not being innately spatialized in a stereo or multichannel sense. A good test of this will be if summing the primary and null outputs produces a true omni.. and if that sounds identical across all three choices of pattern.