In the not too distant past, there were poorly designed MPPT controllers that actually did consume more power than they "magicked" by swapping volts for amps. The most famous of these was a BZ controller rated for 30 amps, which under performed compared to PWM. It was competitively priced, too!
I do not believe there is any doubt that MPPT can outperform PWM. The question should be can this be done economically or not.
Once someone maxes out the roof space, then the only way to grab another 8% or so is to go to MPPT. If they have enough solar harvest using a good PWM controller with temperature compensation and adjustable set points, why would they use MPPT?