RMS is a better way than averaging for getting some kind voltage measurement for a sinusoids signal; its closer but not good. A PWM Supplys output is a square wave with changing duty cycle so theres nothing that can accurately measure the voltage, its all artificial. The only thing that can be measured is amps, frequency and duty cycle and even that dependent on how the PID loops are set up for accuracy.
If anyone is interested in measuring the power going out of a PWM supply you measure amps, the number of electrons per second going through your load, but weve all taken the easy out and been talking volts.
We dont even try and measure volts out of a PWM supply if its mission critical for the operation of the load (like a paper motor motor or X-ray tube) we put an encoder at the load to measure rpm (or rad sensor for x-ray) and use that for the feedback to control the voltage of the PWM output. From those inputs and the current feedback we get the power.
Its the same for us, were using our experience to measure the power and its not set up right on these PWM machine, everyone says they hit harder at the low settings, the POWER is off on all of them.