I just came back to this thread after a long departure from it, and reading it all again has provided a forgotten perspective. Has anyone stopped to consider the need for a small amount of rudder input for both sides of the power curve? Increasing and decreasing throttle? Pulling a sharp elevator input at any rate acts like a brake and is the same as a change in power setting, with the appropriate propeller and torque effects.

A symmetrical airfoil generates lift in part based upon incidence angles and angles of attack. If one wing is "higher" than the other in any attitude one wing will generate more lift than the other wing. Without the proper rudder input the plane begins to "skid" opposite the bank. This can happen at any attitude and airspeed and is a part of an "accelerated stall". We need to all not be so lazy with our left thumbs.

Regarding the question about why the rudder is so effective? It's because it has to be for the plane to fly correctly with the positive rudder response we all like so much.