Exactly. Avg torque/hp is the way to evaluate engine power. Peak is not nearly as important. With a typical 2-stroke, dyno results usually aren't the best indicator of in the field performance. Mainly because they only assess WOT and only at a controlled rate of rpm increase.
I've been involved in many in the field vehicle acceleration tests. The low rpm engine response is a key factor in improving acceleration.
A few weeks ago, I loaned a special vaporizing anti-reversion 4 barrel plate to a friend for testing. He gained .2 seconds in the quarter mile.
The car runs in the 9 second range. His previous setup had so much reversion in the carb at WOT that the wide-band O2 sensors were reading around 12:1. When he installed the plate, it went to 16:1. He had to add 4 main jet sizes and it was still 15:1. He ran it anyway. Said it made so much more low end power that he hit the rev limiter before the traps and lost control. This was at Milan Dragway. Had to deploy his chute. Car flipped and was destroyed. He is purchasing the plates for his new car.
I've done the same testing on snowmobiles, dirt bikes, PWCs, etc, with similar results.
The point is that a dyno can't easily assess this effect. DTR Jim and I have discussed this issue numerous times.
Any engine mod that broadens the power band will usually improve in the field performance.
So, don't get hung up on peak power dyno readings.