As far as I recoll my experience with CO2, the laser itself is CW. So the voltage is simply switched on/off, with maximum frequencies in range up to several kHz. But the voltage itself is not determined, normally the working current should be set and the control circuit applies the right voltage to get defined current (and power) in tube. The current setting is actually an analog value, but can be also set digitally (DAC) or by means of PWM. So there are two possible expanations for the attached page:
1) controller sets the average power in each "pulse" by comparing with analogous signal, corresponding to the applied PWM signal. So each working "pulse" is actually a CW with constant power.
2) each "pulse" is actually consits of the few PWM pulses, each one having some predefined (not regulated, probably max) current. So the average current is set, but actually at any given time during the working "pulse" the current is ether zero or max.
The former is more correct, but in practive the latter is much more likely. I put the "pulse" in parentheses since it is formally not a pulsed laser, but is electrically controlled by pulses.