why not forget all the pedal smoothness and cadence misonceptions alltogether?
Just focus on producing the required effort for the required time duration, then, post-ride analysis should take care of what cadence you naturally are using at different power output.
Good idea; that is how I got on to this idea in the first place. But I couldn't really quantify it until I had a power meter.
In the past, timed efforts always had some uncertainty due to weather conditions and day-to-day variations in perceived exertion. And when going by perceived exertion I always choose a higher cadence. It was only by noticing that the power was actually higher when pedaling at very low cadence that I started questioning. With the power meter it doesn't have to be post-ride, because I set the unit to show both 3 second power and power averaged over the interval. So it is instant faeedback with the average also displayed for any length interval I choose.
Then, I also thought that maybe I actually am "exerting" more at high cadence, but it is just inefficient exertion that does not translate into power to the cranks. It does make some logical sense that efficiency would be higher at low cadence, because you can imagine that if taken to extremes, you could flail away at high rpms in such a low gear that you are producing no power. But you can always stomp hard on the pedals even at very low cadence.