What Todd says about angle of attack on the ground makes sense and is
absoulutely correct - the problem is that we are working with wings
well below their stall speeds, with weight supported by the landing
gear, so "angle of attack" takes on a different meaning: at low speeds
and "high angle of attack", the airflow around the wing is probably
highly turbulent, and control surfaces will be inefficient. Reduce
that "angle of attack" and the control surface is working in
undisturbed (or less disturbed) air, and should be more effective.
Raise the tail, and turbulence should be at a minimum.

I would love to see some wind tunnel testing of, say, a Pik-20 wing at
low (below stalling) speeds. I would bet a high tow that at the normal
ground attitude, on the ground (i.e. a ground plane below the wing,
weight supported by the gear), at low speeds (zero to 20 knots?), the
airflow over the wing is probably really turbulent, and that setting
negative flaps/ailerons moves those control surfaces closer to
"undistrubed" flow, giving them a little more effectiveness.

Remeber, this is a short, transitory phase of the development of
airflow and lift over the wing - I wonder if much study of this
"pre-flight" regime has been done.

Then again, I may be full of sh*t!

I do know it works on a Pik-20b, and that my LS6 doesn't really care
where my flaps are during takeoff (but I use negative anyway, and go
full negative on rollout - gets the flaperons out of the rocks!).

Kirk
66