Here's the logic. I don't think it's flawed, may be incomplete.
Sheldon Brown's website (a road biker) was the first place I read about the advantages of lowering pressure to let the tire absorb some of the terrain. Obviously in road bikes, its just minor irregularities. Every time a bike goes up and over an irregularity, it bleeds enough energy to lift the entire weight of the bike by that amount. Energy has to come from somewhere, and it comes from pedaling. And so, while most people ride their bikes with 100+ pounds of air, he recommended 90 or even less and also pointed out that pro riders use considerably less. This is consistent with the New Zealand power meter tests that you provided the link for.
However, it's like race cars where you reduce the unsprung weight as much as possible (partly for different reasons). On a hardtail, all of the rear weight (including the rider) is unsprung. But, if all that has to move to absorb the bump is the suspension, then it's not nearly as much work to lift the rear suspension as to lift the whole bike.
MAYBE. But GEAX also didn't support their recommendations with power meters or any evidence at all. And none of this even talks about relative grip. My son is a off-road Jeep fanatic, and they do the same thing. They air the tires down to 5-10 psi for grip off-road and over rocks even if it requires more horsepower to turn them, and then air them back up for the ride home. Most hard core jeeps have an aftermarket air compressor under the hood just to air the tires back up for the road.
I would guess that long MTB endurance races run higher pressures than shorter races???