I have learned quite a bit as a result of this. Irony of ironies is that when I was getting these results in terms of my measurement, I assumed it was some sort of in-room response rather than an issue with the crossover. My crossover designer (who shall remain un-named) did a nice job on my first two products. Frequency response was flatter (see attached measurements from Excelsior below) and the consumer and critical response has been almost universally positive. I have had other speaker designers pull me aside and heap praise on those two speakers.
Regarding measurement: Yes, these flaws showed up and I ignored both the measurements and my ears because I was a fool and believed the relative infallibility of my CO designer. This was driven by how good the first two speakers turned out and by oddities elsewhere in the frequency response curve that I was certain were oddities in my room. I had a big dip in measured response at 120hz for instance that was driven by the room. I have since been given some solid advice how to isolate the room in terms of test measures which I will be implementing on all future designs. I will trust my microphone moving forward.
Secondarily, I made changes in both my bench-marking process and placement which impacted the resulting design. These were benchmarked, not vs my external competitor but my own, higher end speakers which was a mistake. I accepted that since these were less expensive I should hear a degradation in terms of sound profile rather than focusing on my competitive benchmark and ensuring these outperformed it. I also know that my placement was closer to the wall in terms of voicing which accentuated the bass. For the MG 1, I also noticed and intentionally turned the speakers off axis during the voicing process because the frequency response was smoother. This shows up in the data.
These are mistakes that I shouldn't have made and these are clearly rookie design mistakes. They will not be repeated. I will trust my software, always use a competitive benchmark and set a more rigorous set of QC controls during the voicing process. This was sloppy.
All that being said, neither of these speakers sounds bad and in-room, I think the AL 1s sound pretty good. The measurements look don't look great but in terms of the actual listening experience, they can be quite pleasant with proper setup. Setup may be a-typical (near the wall in both cases and not toed or minimally toed in the case of the MG 1) but you wouldn't walk into a room and be horrified. Don't get me wrong, for the prices I need to charge, "not bad" is unacceptable. I need them to sound great and that is and always will be my mission.
One note, the slightly accentuated treble you are seeing in all of my speaker design is driven by the fact that I like a bit of top end sparkle. That is definitely me and my taste and I would not change my Nightshade or Blackthorn models regardless of what measurements say.