That difference is more area under the rpm curve for screen door part control of clamp pressure vs centrifugal application of clamp pressure. It's the same reason the TF/FC guys use timers to limit centrifugal assist at the throwout bearing early in the run. The advantage is basically that controlled slipping of the clutch nets a higher average rpm, which allows the engine to produce more power strokes over the duration of the run...more area under the rpm trace.
Centrifugal control of clamp pressure forces you to start below the clutch's lockup rpm and "drive into it" to be effective. That's where SoftLoc gets it's name. What they don't tell you is that after you launch, that centrifugal application of clamp pressure after the shifts is anything but soft. Just look at your rpm traces after you make a WOT shift- that SoftLoc locks up so hard that it jerks your rpm almost straight down, until rpm is pulled down far enough for the clutch to start slipping again. The screen door parts method does not add clamp pressure with rpm, so the clutch can begin meaningful slipping after a WOT shift as soon as you pull your foot off the clutch pedal. The result is that rpm drop after the shift is more diagonal on the graph, which adds area under the rpm trace.
NMRA classes such as Coyote Stock are required to use spec crate engines, spec transmission & ratios, and diaphragm PP's, an attempt to keep the playing field more level between the haves and the have nots. Ace came out with a centrifugal assist diaphragm PP which is legal there, but skirts the intent of those cost control rules. Guys that stepped up were used to shimming their diaphragm PP's nearly every run to keep clamp load in it's sweet spot. Centrifugal assist looked like a way to greatly reduce clutch maintenance, didn't take them long to figure out that centrifugal assist lowered their average rpm vs shimming.
