As to why the high harmonic distortion around 100Hz occurs, I have a "theory" (rather speculation) to offer.
Have only skimmed the thread do not know if my "theory" has already been mentioned - sorry if the answer is yes.
The lateral slots are nothing else than Helmholtz resonators. These are used with the midrange driver to generate phase-inverted sound pressure and thereby produce the cardioid radiation.
The phase-inverted sound must cover a wide frequency range, which means that the tuning frequency of the slots must be relatively high.
You could also use differently tuned slots to extend the frequency range, but don't ask me what exactly happens then. This and my "theory" would all have to be checked and verified at least by simulations.
It would have been interesting if
@hardisj had also taken near-field measurements of the lateral slots, that would facilitate a possible explanation.
Since the subwoofers have a crossover frequency of 100Hz, the midrange driver still plays at full sound pressure level below the tuning frequency of the slots. However, this means that the excursion of the midrange driver increases extremely and it therefore generates more HD3.
If then at around 60Hz mainly the two subwoofers generate the sound pressure, the distortion drops to an intermediate minimum.