Let's examine that powercrank graph -
http://www.powercranks.com/assets/images/PowerCranks pedal forces.jpg
to see what it shows....
I assume that the testing was done in the same session and with the rider maintaining a constant 250W at 80rpm.
NOT that the rider was attempting to produce maximum power or rpm.
1) With PC, the rider's leg is always 'working' - except for a brief moment at about 315 degrees.
With non-PC, the leg is 'resting' (being raised by the other leg) for about 150 degrees.
2) Without PC, the downstroke torque needed to produce 250W appears to be about 20+ percent more than when using PC.
It would be interesting to see a graph of the rider's torque with non-PC prior to training with the PC.
A 'good test' of PC should include: power output, torque graph, oxygen usage, and pulse rate on non-PC BEFORE the start of PC training.
The test subjects should be strong active cyclists who have stopped improving (so the benefit of PC training is the only additional factor).
Jay Kosta
Endwell NY USA