This is where the misunderstanding comes in. Stabs do not have to life downwards to provide stable flight. The idea that they do is largely from full sized aircraft where the designed pitch stability margin is so high that the aircraft run CG positions in the 25% and forward region and thus mostly operate in a negative stab lift situation. But most model designs are set up for far more relaxed pitch stability factors and positive lifting stabs with our models are much more common even in sport flying RC designs. Positive stab lift is not just for free flight designs.

Mark Drela posted up a lovely chart a while back that shows the relationship of the CG position and "stabilizer" lift for a continuous transition from designs with small stabilizers where the lift must be negative to large stabilizers where the lift is positive and on to tandem wings and finally ending with canards where both surfaces are lifting positively (obviously). Only the one extreme end where the stabilizer is small is the stabilizer lift negative.

Of course this is all based on if you choose to operate any given model at a point close to the neutral stability point. If you operate any of the "conventional" setups with a CG at or ahead of the 25% point then you ensure that the stab must operate with negative lift.

Carrying on.... Any aircraft with positive pitch stability is a single speed design. Any displacement from S&L will cause the speed to change and the balance between wing and stab lift components acting against the CG will change in such a way as to return the model to it's trim speed and S&L flight.

I've read the bit about the airfoil on the stabs supposedly providing an "autopilot" like function but in reality this doesn't happen. What IS at play is the balance between the wing's lift,leverage and pitching moment and the stabilizer's lift and leverage all in a balance. However to avoid large speed related changes in the flight path the free flight models are trimmed to operate VERY close to the neutral stability point. However they also use rather extreme airfoils and it's not uncommon for the various non linearities in lift and structural flex to gang up and cause a tuck and final death dive if the model wanders out of an often rather narrow envelope where the forces are in a positive stability balance. This can happen where the lift of the tail rises faster than the lift of the wing and the lift lines hit a crossover point and the tail lifts the model into a dive rather than the wing lifting it out of the dive. Then it's lawn dart time....

Contrary to what the old time designers may try to tell you the successful designs came about mostly through trial and error... a LOT of error. I know because I've also had my own share of errors when trying to walk the tightrope of this aspect.

The most recent one was a P30 rubber model design where I used a highly cambered airfoil for the wing and a fairly highly cambered one for the stab. If it was just climb and glide then all was well. But if the nose went down due to some turbulence it was a guess if it would pull out or if it was lawn dart time. I fixed this by making a new stabilizer that had a lower camber airfoil. But as a result I was able to operate with an even further back CG than with the higher cambered stab airfoil. Re-read this bold bit again if it isn't clear. It's a large part of the secret of understanding this whole issue.