Powered speakers show audiophiles are confused

OP, looking carefully at the Stereophile measurement plots I think the issue is that the H and V windows are entirely different.

Stereophile uses +- 90 degrees horizontally but a +- 15 degree window for vertical measurements.

Under these conditions the natural tweeter roll off at 30 or more degrees won’t be visible in the vertical plots.  If the GR research video uses the same windows then the same issues will apply.

Normally in the vertical domain speakers perform better in the frequency domain with higher order filters because there's less destructive interference. 

@erik_squires - what time mark in the video. I looked at a few plots on another video and noted strange off axis.

@thespeakerdude Sorry my friend, I don’t care enough about this video to go looking again. If I assume he’s using the same differences Stereophile uses in charting off-axis response then his plots make a lot more sense.

The dip at the crossover region in the vertical plots is quite common, though I seem to remember a Tekton stand mount with absolutely horrible vertical response plots.   If anyone is interested in comparing vertical FR with different speakers.

@erik_squires it seems odd to me that speakers don't do frequency plots like microphone plots showing off axis coloration at different frequencies. The importance is just the same, with microphones generally the longer the polar pattern the poorer the off axis colorations are and that is very important to the overall sound depending on the acoustics of the room and where there are other off axis noises. It is the same for speakers poor off axis coloration that bounces off a wall is just the same as a microphone that hears poorly from the side. 

@erik_squires I feel the same about the video :-).  If it's the dip I think you mean that is likely just the stacked two way dip.