Reynolds number effects might get into it ????????

I honestly have no idea, but something is nagging me at the back of wherever those things that nag you come from. The thing about being old (well 60 and not that old although badly beat up physically) and senile (ask my lovely wife) is that things seem to hover just out of mental grasp.

As far as loads go if there is sufficient dynamic pressure over the surface to get airplane response then the tab might also be working for you. The higher the q the better it works. In a low q condition all you are moving is the weight/inertia of the control surface, not a lot.

That being said, if I were making a big TOC machine I would build in trim tabs and stick a medium load servo on them. Unloading the main servos is not going to hurt anything that I can think of even if all the loads are not being removed. As a test you could start out with all four on rudder, then unlink one at a time until some lack of rudder response or precieved deflection rate is noted. As was noted you could end up with several hundred dollars of savings.