I can't see how it would make a noticeable difference.  Relatively speaking, things just don't happen that fast in control line.  A typical PWM signal updates every .02 seconds.  For a typical stunter that equates to about 16 inches of linear travel and approximately 3 rotations of the prop.  How much change in velocity happens over that 16 inches?  Is it enough to warrant faster updates?  Oneshot125 can update 8 times faster, approximately every 2 inches along the flight path and 1/3 of a rotation of the prop (maybe 5 pulses from the RPM sensor).  What is the loop time on your PID controller?  How accurate is the RPM calculations based on fewer pulses?  A lot of things to consider.

I think you would see a much larger gain in performance just by using ESC's that have very little to no input filtering before going to a faster protocol.

CF,
Are you using 8 as the PWM in the ESC for your .02 sec? The BadAss motors use 8 as the standard starting point. I guess another option in to increase that to 12, but this will increase temps in the ESC and motor some need to make sure the cooling is very good.

Best,    DennisT

CF,
Does the PWM rate set to a higher frequency (say 12 or 16) do anything for quicker response to load (like in a hard conner and vertical climb?

Best,    DennisT