Hmmm, I have to agree with Paul...

The part of the discussion that I do agree with, is that by placing the pivot at 25% MAC, and assuming the airfoil is symmetrical, the torque required from the servo would be very small (theoretically it is zero, but with friction, inaccuracies and other real-world effects there will usually be a slight load). From an "aeroelastic" point of view (not really the right term since we are not really talking about structural elasticity) - the control surface is neutrally stable. That means you can put it at any angle relative to the flow and the moment remains essentially zero.

If you use a more forward pivot point, you are erring on the safe side. Now you require more load from the servo to pitch the surface, but the aerodynamic load on the surface is trying to move the surface back to neutral. Essentially, if you disconnected the pushrod the control surface will point itself into the flow.

Moving the pivot point anywhere behind 25% is extremely risky. Basically, whenever the horizontal tail produces some load, the moment on the pivot point is in a direction that will want to increase the pivot angle. Bigger pivot angle increases the aerodynamic load, which increases the moment, which wants to pivot it even more and so on. In other words, for any non-zero load the servo will be fighting the control surface from running away, or if you disconnected the pushrod under any flight condition the surface would slam to either the full-up or full-down stop.
Although the servo may be strong enough to counter this constant load at low speed, near zero tail load the surface will want to oscillate, which in turn could develop into a flutter or at the very least make the airplane difficult to control. At higher speeds the servo may not be able to cope with the load anymore and instead of being forced back towards neutral, it will be forced to full up or full down elevator!

The best place to put the pivot point is just slightly ahead of the 25% MAC.