Furthermore, as the duty cycle changes with frequency, the RMS current drawn will increase with the higher clock speed.

Is electrical power a function of its frequency?....
R.M.S current drawn, possibly from the d.c. supply?....
Try feeding a PWM clock signal into a JK f/f in toggle mode, then look with your scope at the duty cycle comming out of Q and /Q, the last time I looked, the duty cycle was 50% exactly, secondly, as I increased the frequency... the supply current does not go up or down, if it did then surely there must be same law that says it should, discarding hf performance efficiencies that are a different matter.
Logic chips are getting faster year on year, but the supply current consumption appears to be getting less and less, i.e., effeciency is getting better, how would you account for this?....

Darren