Here's the way I see it:
Speaking AS an owner of the PDX1.1000, I was AMAZED to find that it does, in fact, run cooler at a 2-ohm load, than it does at a 4-ohm load! After accidentally discovering this, I went to my garage, and (after some searching...) grabbed my trusty 100:1 current transformer, ran the positive speaker wire through it, and proceeded to run some 50Hz sine waves to the woofers I have. First candidate was the stuffed TL in the trunk, with two 10" iso-loaded 4-ohm DVC TCSounds TC-1000 woofers. This was the 4-ohm setup. The second, my new Eclipse 10" Ti woofer, wired in parallel for 2-ohms.
All of the current measurements worked out - without changing ANYTHING on the deck, I got somewhere about 16A into the 4-ohm load, and 23A into the 2-ohm load. These currents correspond to somewhere in the 1050 watts range, a bit short of the 1176 watts the amplifier was birth-sheeted for, but I really didn't notice too much of a rise in current beyond this.
All I can think is that the power supply raises it's PWM FREQUENCY, to optimize the output rails for the attached load. Toroid transformers are very efficient in power conversion, and become even moreso as the input frequncy rises, especially relating to current conversion, meaning they transfer more energy OUT of them for a given energy INTO them, as the PWM driving frequency increases. I've opened this amplifier, and the
toroid in it is TINY, (see the red ones on the left of this photo) compared to some others I have seen in this power category. Due to this, something MUST be "smart" in there, else I would never be able to observe this type of behavior.
While doing this, the power supply stage SPECIFICALLY works more efficiently, reducing the thermal output. (As a side note, MOST of the heat generated in any amplifier is in the DC-DC stage, anyway.) The OUTPUT stages in a high-quality Class D amplifier approach 95+%, and don't really change much... Whereas the OVERALL efficiency of a Class D amplifier, (all stages included) are around 80%. I've felt 150watt DC powered (as in high-voltage primary battery - really DC rails) Tri-Path (RIP

) Class D amplifier modules, running at full output and little more than heat-sink modules big enough to cover the module, and less than 1/4 inch thick, and they haven't EVER felt warmer than ambient.
In this case, THIS amplifier will only do 1000 watts into EITHER a 4-ohm load or a 2-ohm load.
If I'm wrong with my "rememberies", PLEASE
i am an idiot, call me on it! You're the techie that all thechies *I* know aspire to be when they grow up!
It all reminds me of something that Molière once said to Guy de Maupassant at a café in Vienna: "That's nice. You should write it down."