Sure work and energy are related, but in this case most of the time the servo is stationary, either at rest or deflected some amount so the majority of the time we are concerned with the static load on the servo which is not work (no movement), but translates into electrical current.

Take the rudder for example. If we're flying straight and the rudder is at neutral, the load on the servo is little. If we are slipping in for a landing the rudder is deflected into the airstream during the slip. The air rushing by imposes a load on the rudder which is taken up by the servo (torque). This load becomes current. If you move the control rod at the rudder further out on the control horn it will take less servo torque to hold that same displacement, thus less current. It will also take more servo travel to reach the same rudder displacement, but that additional movement is a secondary effect.

You should always try to maximize servo travel during setup, that does not mean that you should maximize control surface deflection. By maximizing travel you will also maximize resolution, another good thing.