Originally Posted by Charlie
It all depends who you beleive.
Some would venture to say the power cap is there all for "ripple current".
When an ESC switches on and off, the power wants to keep going when the switch is off. This creates heat and slowly hurts performance. The power cap catches the ripple current when the mosfets are off, then gives it back when they switch on.
In turn, you get more voltage to the motor that would essentially be lost as heat.
There are "lots" of ways to look at the side effects or how it feels or what happens. I love the internet because you can read ever version twisted one way or other, and in some respects "everyone is right" a little.
I spoke with charlie and bob about this subject shortly before I departed the race hotel for the Novak race.
Looking at power supply designs for other applications, I noticed that the function of a cap in an esc is much like an LCR filter. Bob and charlie confirmed this (as charlie posted above). Speed controls output PWM which and the cap makes this occilation less dramatic, and as charlie mentioned can cause the high end of the pwm cycle to actually be at a higher voltage. To see a more dramatic version of this effect look at a cockroft walton charge pump which uses an AC or pwm source with a network of caps and diodes. Basically the pulses stack up in the caps increasing the output voltage.
The whole thing about the cap unloading out of corners isn't really why they make the esc punchier, that would barely crank the motor over for half a rotation.