The comparison is made with the same controller and the rest of the set-up the same so we don't need to speculate about how a different controller might do instead. I don't know if the controller's temperature compensation (internal, not remote to battery) matters. AFAIK that just affects the set-points. The controller is in the shade in a compartment with the door open lots of air.
I have used PWM set-ups in the same location at these same high temperatures with no loss of amps. The 12v panel does also have lower voltage in the heat (20.3 vs 21.9 rated typically) but the Isc actually rises (positive coefficient) and I get Isc as the output.
So the comparison with PWM is under the same conditions. What I find surprising is how the MPPT amps are less with just that much heat (24C ambient). I had the idea you had to be in Death Valley or someplace to really notice that loss of amps.