For Bingo possible bull!!!! alert
Okay, I'll give it a first try. For brevities sake, let's assume the (fixed pitch) prop doesn't have any twist but has a uniform 18 degrees or so AoA. (With AoA in this respect I mean the AoA of the prop blad when turning, with the aircraft stationary. So this AoA will decrease once the aircraft starts speeding up.)
In level flight with the prop stopped this means the AoA is now 90-18=72 degrees, well above the stalled angle. Now you're going to "pull some G's" to change the AoA of the prop by changing the AoA of the whole airframe. But you can only change the AoA of the airframe up to the airframe stall AoA, either positive or negative. So that's roughly again 18 degrees one way or another. Best you can do to the prop AoA is therefore 72-18 = 54 degrees. Still well above the stall AoA.
Now of course the AoA of the prop is not a uniform value throughout its length, due to the twist. Near the root it may be as much as 45 degrees. So at that part of the blade it *may* just work. But that area is also the place where the prop is normally least effective, plus (due to its proximity to the root) the required moment is also the greatest.
And anyway, between the 45 degrees at the root and the 18 degrees at the tip, the prop is still stalled throughout its length in level flight, or in a level dive. And still, in a dive with sufficient airspeed the prop will start windmilling anyway. So I don't think stalling/unstalling has anything to do with this. After all, even a stalled airfoil will create some (impact) lift.
Nevertheless, now that I'm thinking about it, I think the whole idea of changing the AoA of the airframe has some merit, but for a different reason.
Suppose you've got a prop that stops in flight in a more or less horizontal position. And you've got a clockwise-turning engine (as seen from the cockpit). In level flight both blades present a more-or-less flat surface to the airflow and nothing much happens. Until the impact lift against the not-quite-perpendicular blades is so strong that the prop moves the engine through the first compression. But anyway the drag on both sides is the same.
Now you're changing the AoA of the airframe by pulling Gs. This means that the prop blades now have a different AoA too. The blade at the left hand side of the airframe presents more of a leading-edge-on profile to the airflow, so its drag is lowered. While the right hand blade presents more of a perpendicular profile to the airflow, so its drag is increased. It might be this difference in drag that helps push the engine through its first compression.
OTOH, I apparently failed the QFI course (though I always thought you couldn't fail if you didn't try) so I'm standing by for a better explanation...
