Re the material in post 50:
Yes, as has been pointed out many times, as crank inertial load increases, I.e. at speed on the flat, it becomes more efficient to hammer the downstroke. People like Dylan are supposed to know that, but some folks just like to push one viewpoint 'cause it's simpler. To see a very different force application, one can look at climbers, Froome for example:
Yup, he was a heel dropper while climbing, even at high cadence. Climbing, even as fast as these riders go, is a lower crank inertial load environment, hence it makes sense to apply force through a greater range of pedal arc.
I should also point out that if one is as fast as a pro on a 40k TT, that really increases that inertial load and optimizes the downstroke. OTOH, if one is riding at 20kph rather than 40+ kph, your inertial load will be a lot less and your optimal pedal stroke will be different.
Another thing to think about is endurance. It seems to me that more one can spread the pedal stroke out over more muscles, the less each muscle has to work and the longer they'll last. I've focused a lot on my pedal stroke over the past 25 years and found that it is certainly trainable and also has a big effect on my performance, enabling me to ride above what my age and FTP predict.
Riding around here, it's all about climbing speed, and thus concerns pedaling form. My little oddity is that I do all my indoor training on a set of resistance rollers. My bike has light tires and tubes and the rollers themselves are ABS, so the whole arrangement has a very low inertial load. That’s the reason that they say rollers make you smooth. So I’ve always focused on smooth pedaling, which has served me very well on local climbs.
Toe pointing or heel dropping is sort of another issue. I'm a heel dropper because it puts less strain on my calves and so is easier to do for very long periods.