I will throw in a flat spin experience many years ago. It was a moderately powered .15 glow shoulder wing model. It flew nicely and looped and rolled and could spin nicely. The recovery was OK and not scarey at all. The airplane had a rather small vertical tail but it flew OK. I decided to add a canopy and pilot over the center of the wing.

The next flight I dropped nto a spin and it went from normal into a really flat spin. The descent rate was so slow I was able to try all variations of power, control throw, begging and praying and all were to no avail. Remember this was a moderately powered airplane with small surfaces that couldn't be powered out of the spin. The true flat spin was also maintained with no control inputs.

It came on down and gently settled on the grass. Pretty nice landing if rotating counts. I took off the canopy and all returned to normal.

There are a lot of 3D designs that can be pushed into and out of spins, flat spins, flat spins going horizontal, etc. but the pure flat spin that is based on the combination of inertias, CG, configuration and corresponding aero inputs of a non-augmented kind (no propwash across deflected surfaces) has a spin boundary that is defined by a ratio of damping to forcing parameters and can be stabilized by several combinations of the mess (my aero friend in the next desk to mine did the F-15 test, spin evaluation and report, and tried to explain it to me as I had to proof read it, but there were times I got lost in the math model).

Trainers may be very resistant to entering a spin but if that combination hits, it will drop into one, if not the combination, then never. It would be fun with a foamy for something like a basic Extra design to run a series of experiments with CG and vertical tail/lateral areas and see where the magic location/areas are. Too scarey to do with a real model.

Dick is a good example of if you aren't good - be lucky. ......... did I word that right?