Downthrottling was part of preparing for combat from CRUISING speed... Slow turning from the start was his combat speed all the way through most of what the enemy did, except diving perhaps...
You hit the brakes in a corner, but not all the way around the racetrack...
Said one of the P-38 pilots: "He seemed throughout the ordeal to be loafing about at reduced power with his canopy open, always turning with the needed angle to spoil our attacks..."
Regarding the top excerpt, consider the following:
1. your assertion about the disk loading states that the effect of the reaction torque reduction caused by downthrotting is immediate. I accept this argument, though not your scaling argument (see HT's estimate and my own response to his estimate in which I, in turn, estimate the relative velocity and alpha change on the blades due to turning). Why, then, would F-dub pilots feel any need to downthrottle
prior (better answer: to get to something like the right speed for max dps/min radius -which would involve a delay, turn onset, then upthrottling to overcome the increase in induced drag) to combat? Further, if the practice of reducing power to improve turn rad and rate were used on 190s, why would it not also make sense to use the same effect on other aircraft - all of which would experience the same effect? Yet, it is not tactical practice to do so on any other notable type... Were the pilots of all other types presumably unaware of the effect?
No, your account doesn't sufficiently explain why the F-dub would be the only type to be flown this way. Badboy's explanation, otoh, is far more plausible.
2. At the same time, I find it entirely possible that wanting to go into the fight at something more like corner might make more sense for the pilot than entering the turn at a higher speed and allowing decel. Why? Because prior to decel and at pilot limit, your dps and rad are both worse than at corner - i.e., for the same g load, at the higher speed your turn is larger (again, g=v^2/r). I'm also curious, at this point, about the relative acceleration performance of the FW as compared to, e.g., a Spit VIII or IX. If the Spitter holds E better (and I think it does), the F-dub might, by getting to corner prior to fight start, have a brief turn advantage over the Spitter. Consider, for example, a Spit that can make a 4g pilot limit at a lower speed than the F-dub... if it's currently at 325mph and the F-dub is already at corner, who has the turn advantage? - given that the pilot limits are similar, the answer is something like a 9/4 ratio in favor of the a/c at corner). I'd also cautiously here assert, if indeed th e Spit is worse at accel/decel that inducing an overshoot or "too-large" energy scrub from a Spit-flying foe might be a decent tactical idea for the F-dub jockey.
3. Regarding the Ki43 and P-38s, I have little doubt that a good Ki pilot could make decent evasives of attacking high-speed 38's from a reduced power setting. This says nothing objective about the impact of power available on sustained turn rate.
4. Combat Flaps: we don't have 'em on the F-dub in AH but you could expect their effect to reduce your corner speed and increase your best turn rate/decrease turn r.
For a decent development of this, I might point you to BadBoy's fairly neat ACM guide (kudos on this, BB):
http://www.netaces.org/badboy/Acmi.pdfSo, I see little reason that popping flaps is something too difficult to understand IF we accept that the FW pilots were behaving as posited above in 2..
5. You DON'T hit the brakes in a corner (unless you want to spin). The cardinal rule of racing is "throttle in a corner, brake in a straight line". This simple statement betrays a misunderstanding of the weight transfer impact on traction in an auto AND a misunderstanding of the fact that more power is required, both in flight and in an auto, to overcome, in the case of the ac, greater induced drag caused by "uplifting" or, in the case of the auto, the greater rolling resistance caused by maintaining a slip angle to generate cornering force.
In short, I think the FW argument you posit, while interesting, is caught in a logical trap (i.e., why is the practice limited to this type? your explanation is insufficient), a theoretical trap - see any decent development of the turn performance min radius equation, and an empirical trap - see test data on just about any type of aircraft for which turn performance data is available.
Anyway, that's my read on this thing.