To tie in Nick's comments and the Vsquared perspective, I would suggest that, in the example - excessive power pedal (left pedal in American helos) - the increasing torque requirement from a reduction in RRPM is caused by Vsquared.
As Rotor RPM is bled off, collective is increased to maintain a constant hover height. The reason this has to occur is because of Vsquared in the lift formula:
Lift = Coefficient of lift X 0.5roe X Vsquared X S.
Therefore, as RRPM decreases, Velocity over the blade decreases, and this effects the lift exponentially because V is squared.
As you need to produce the same amount of lift to stay at a constant hover height, the only variable to change is the coeffecient of lift - ie pull more pitch to increase Angle of Attack (AoA) thus uncreasing drag as well, hence requiring more torque. Clear as mud?
As Nick says, the extra torque would require more left pedal, thus helping your situation. But this is not all that helps. As the tail rotor spins at a fixed ratio to the main rotor, but at much greater RPM, a decrease in main rotor RPM will result in a proportionately larger decrease in tail rotor RPM, and hence V over the tail rotor. As V is squared, lots of lift is lost from the tail rotor, and becuase the pitch is fixed, total tail rotor thrust is reduced thus helping you even more. Going back to the UH-1H, the tail rotor is nearly always at it's peak in the hover so any reduction in main rotor RPM really reduces tail rotor thrust quickly, and small main RRPM decreases are very effective in stuck left pedal forward cases. Incidently, this quirk has caused/contributed to many limited power accidents in the UH-1H, and I seem to recall some mention in the FM about it. Can any one recall?
I guess it gets more complex on aircraft equipped with mixing units because the effectiveness of rotor bleeding depends on whether the pedal jam is before or after the mixing unit. I am guessing that is why it is a hard procedure to practice on these aircraft. Nick?