Trim Drag, Static Stability, Stick-free stability
Hi, Lets see if I can help here:
First, to put your mind at ease, there is very very little drag caused by trimming with small elevator deflection, except for the minor drag of the elevator horns sticking up in the breeze. Even that is pretty small. You can help that of course by putting nice round edges on the horns.
Trim drag, as a term, more classically applies to the additional "induced drag" caused by trimming a stable airplane, not the added profile drag --- although I suppose in a case where there was a really significant drag addition from balance horns or something, it might be included as well.
The extra induced drag is a double whammy - not only do you get induced drag from the horizontal tail as a lifting surface (lifting up or down) but when it is lifting down, the wing has to produce more lift to offset it, so the wing also has an increase in induced drag.
Static Stability: Probably the most widely held misconception in aviation is that the tail must have downward lift to make the airplane stable, and that zero lift on the tail would correspond to neutral stability. This is just simply not true. What IS true is that for a statically stable airplane, the tail can never lift upward at a higher lift coefficient (lift/unit area) than the wing, and including the effects of wing camber, a trimmed, stable airplane may only have an upward lift coefficient that is a fair bit less than the wing lift coefficient. BUT, it is often true that at low speeds, many airplanes have small amount of upward lift if the c.g. is toward the aft third of the c.g. envelope. This is especially true of sailplanes where performance is so important that we always balance the airplane in the aft third of the c.g. range, and because with a short wing chord, the effects of wing camber are lessened, allowing more tail upload. Anyone that doesn't believe this, I would invite you to get any good text on aircraft stability and control and work through the equations yourself.
As a stable airplane is trimmed at higher speeds, more download on the tail is needed. A stable airplane could easily trim with upload in slow-flight or best glide speed, but at cruise speed when the wing lift coefficient is very low (often 0.25 or less for GA airplanes) the tail is most always lifting downward. The faster you go, the more download will be required to trim. It may seem counter-intuitive that it takes a more forward stick position, and a trailing-edge-down elevator deflection, but this is overpowered by the nose-low angle of attack and the wing downwash angle on the tail. Considering only static stability for a moment, the incidence of the stabilizer has no effect on (stick-fixed) static stability. A change in stabilizer incidence will require a different elevator deflection angle to trim. And ignoring other factors, you might choose a stabilizer incidence that will lead to the elevator being in perfect trail at your typical cruise speed. That would eliminate any extra drag from elevator horns. But, as the engineer from Vans (Ken?) was quoted as saying, there may be good reasons not to do that, because:
Stick-force gradient/stick-free stability: For the case of RV's, where the elevator airfoil is symmetrical, if you have too much stabilizer incidence, you will find that the stick force gradients are very light, or even negative. It has often been commented that RV-8's, near the aft c.g. limit, have very very light stick forces at low speed. The CAFE flight test even shows a small negative stick force gradient at low speed with the c.g. on the aft limit. Those of us that fly high-performance sailplanes are accustomed to very light stick forces, and thats usually OK, but negative stick-force gradient is very undesirable. It means that as you slow down, you have to provide forward stick pressure to prevent it from slowing further. If you slow more, then more forward pressure is required. I've seen an extreme case of a badly rigged home-built glider where it also took more and more aft stick pressure the faster you went. Really unpleasant. This trait is often interpreted by pilots as being statically unstable, but it is not. It is, however, stick-free unstable.
A number of factors in the control system design and aerodynamic design influence the elevator hinge moment that feeds through to the stick, and how the hinge moment changes with elevator deflection and speed. But having too high a stabilizer incidence angle is one of the factors.
So my recommendation would be for most RV's, build it according to the plans. This is one of the things that the factory may have played with on prototypes to get to where they think its right, or they got lucky and hit it right the first time. But in general, part of the RV's great handling and feel is because they got the incidence right. If you are going to err, do it on the side of less elevator incidence (err in the direction of leading edge down). This will provide more desirable stick force gradient - what most pilots would associate with a more solid feel. The change in performance (trim drag) from adding some (L.E. up) stabilizer incidence would be really negligible, but the change in handling is noticable.
If you are building an RV-8 and you know you will often have a heavy person in the back seat, it wouldn't hurt to drop the incidence just a tad (something like 0.1 or 0.2 degrees L.E. down from the spec), but whatever you do, don't err on the side of more positive incidence.