Yup, it works in my flying boat. It’s good for a couple of MPH, but can be hard to find sometimes. If I’m really heavy, it may not be possible, but lighter, it works. I do agree that some airframes respond better to this than others.
What I’d like to know is how this works. I cannot see how it can work except in the scenarios mentioned i.e.
If there is a difference to be observed (of which I’m still skeptical), might it have to do with the behavior of the prop governor on planes with a constant speed prop, together with the engine? Maybe something to do with “flying” the prop before or behind the power curve depending on if you come from an “overspeed” or “underspeed” situation?
At piston engine GA levels the aircraft is operating much more to the top right of Peter’s Power/Airspeed graph, whereas at high altitude you can be much closer to the bottom of the curve, where there genuinely is the possibility of 2 different speeds for the same power setting.
I don’t think the power plant or propeller has anything to do with it, it’s to do with the much higher TAS (or lower IAS) at high altitude.
If the power-required curve were the whole story then the aircraft would of course always eventually accelerate to the same cruise speed, where excess power was zero.
But the power-required curve is an approximation, it only deals with airframe drag variation with speed, and there will be other secondary effects at play.
For example, a few knots below the cruise speed, alpha is higher, and while this is reflected in the power required curve in terms of parasitic+induced drag, there will also be increased P factor and increased rudder (trim drag) to offset the P factor. This trim drag is not reflected in a simple power required curve, and may be enough to sap the few HP of excess power that you would have otherwise used to accelerate.
In Jason’s example, I suppose that engine efficiency may be lower at the lower speed, so while in theory the engine is at X% cruise power setting already, a temporary excess may be required to accelerate past that regime?
I’m sure there are other secondary effects, they will all be small, but can in theory still “rob” your excess power and require either a descent or a power increase to accelerate through them.
If I read you right, Ortac, you are suggesting that the classis pitch control mechanism (the elevator and elevator trim, and the location of the CofG relative to the centre of lift) has some sort of hysteresis in it even if not flying anywhere near the back of the curve.
I"m just saying that the typically seen power-required curve is a simplified model which is a function only of “pure” parasitic drag and induced drag components. With the former being directly proportional to the airspeed cubed. And that secondary affects alter this curve in the real world.
I don’t think that stability (CofG vs Center of Pressure) relates directly to this, but it could be another secondary effect, as centre of pressure moves with angle of attack, and a different centre of pressure means a different amount of trim drag.
Maybe there is more at play when approaching the absolute ceiling altitude, or even for a turbine engine due to some bootstrap effect….but as the video demonstrates for a piston aircraft, well within its operating altitude at least, where >99% of us operate, there is no noticeable effect…
Now, about those downwind turns….
I did find this Rogers paper which may have been the “polar” which Neil was thinking of?
That paper appears to confirm that the concept exists only at high altitude or slow flight:
" Thus, in practice, only at the higher altitudes does the ‘step’ exist"
IOW, it exists in the region where you have two possible speeds at the same power setting.
That’s unless there is some other subtlety which I don’t get, which is quite possible 
