If we ignore turns at very high speeds that are less sustainable for most WWII fighters, say above 320 MPH, where the P-51 would shine, it is actually fairly easy to make a sustained turn rate hierarchy at medium-low speeds, and one that agrees with most tests and seemingly contradictory anecdotes.
If you take the A6M5 Zero comparative tests as a benchmark (from Mike William's "WWII Aircraft Performance" site), all the U.S. fighters are set in a clear hierarchy.
The A6M5 beats the following U.S. fighters by 360° in the following amounts of sustained turning (roughly the same spiralling up or down);
F6F= in 1260 °
F4U= in 1260°
P-47D-30= in 540° approximately.
P-51D= in 570° approximately.
P-38J-25(same as L)= in 720°.
A U.S. Navy tests FW-190G, fully disassembled and re-built wrong for tests with inferior results= 583°. The FW-190A-4 was tested by the British as being equal in sustained turning to the P-38G, which I think pegs the FW-190A-5 as at least in the class of the P-38J-25, or around the same 700-800° range. There is massive evidence to support this excellent FW-190A low-speed turning performance, and also its poor high speed handling performance. See this Russian evaluation:
http://www.ww2f.com/russia-war/21828-russian-combat-experiences-fw-190-a.html Or this Spitfire ace account:
http://img30.imageshack.us/img30/4716/jjohnsononfw190.jpg In 1200 combat reports I have read on Mike William's site, there is massive evidence to support that the earlier P-47D Razorback is superior in turning radius to the Me-109G in almost all circumstances, if less so to right... German tests of a captured Razorback are unequivocal: "The P-47D out-turns the Me-109G". From: "On Special Missions: KG 200".
Against early to mid FW-190As, such as this test of an untampered-with FW-190A-5, the lowly NON-paddle-blade prop Razorback, as in real combat accounts, is inferior to the FW-190A at low speeds, but would be much closer in my view with a paddle-blade prop, as early '44 combat accounts indicate. Even with the inferior needle-blade prop it is an interesting give-and-take close match:
http://img105.imageshack.us/img105/3950/pag20pl.jpg This in my view would peg the needle-prop Razorback P-47D at 500-600° in low-speed turns, but much higher with a paddle-blade prop, perhaps as high as 700°-800°, which would explain parity with early to mid FW-190As, and the evident superiority to Me-109Gs devoid of MW-50 as the Russian tests say: "The FW-190A is more maneuverable in horizontal flight than the Me-109".
In the P-51 combat reports section, the same level of turn superiority over the Me-109G is not really there, resulting in long turning matches were the P-51 has difficulty finishing off, probably because it has a wider tuning circle that is maintained faster but requires an extra, and risky, angle when firing from the "cold lag" side (outside the turn). Superior speed retention in turns for the P-51 is very evident against the 109G, or even against the FW-190A IF at high speeds, except against later-'44 period MW-50 boosted 109Gs, which give even more trouble to the Mustang...
I would peg non-MW-50 Me-109Gs in the same 550° class as the Merlin P-51 versus the Zero 52, but the 109 was indeed much better with MW-50: 600 to 800° is possible, which would then surpass slightly early to mid FW-190As. Note that MW-50 required C-3 fuel, and this became a priority later for the FW-190As which could not use anything else: I think MW-50 in 109G/Ks was troublesome, and not as common as is often assumed...
Note also that the FW-190A-8 was widely known among German pilots as a massive improvement in low speed turn performance over earlier 190As, especially with the broad wood prop, and vastly out-turned the less agile bubble-top P-47D (very evident in late '44 combat reports!) that was itself roughly equal to the P-51D. One reliable account from an actual FW-190A-8 ace has the 190A-8 beating the P-51D, shuddering once or twice at the edge, in two right 360° turns to reverse a tail position. This puts the later FW-190A-8 at 1140° versus the Zero A6M5, or slightly below both Navy aircrafts: Quite acceptable.
So the top pack discussed here would be the F6F, F4U and FW-190A-8, all at around or just above 1100° versus the Zero A6M5.
The mid-pack would be the P-38L, P-47D Razorback (paddle-blade), FW-190A-4/5/6 and maybe MW-50 Me-109Gs, all around 700-800° versus the A6M5.
In between that and the bottom maybe the non-MW-50 Me-109G-6?
Then the bottom of the pack would be the Merlin P-51s, Bubbletop P-47Ds and perhaps the gondola-equipped Me-109G-6, all at around 500-570° versus the A6M5... Note these U.S. aircrafts are still the better overall relative turners above 300 MPH...
The air show pilots comments are based on less than full War Emergency power, so at lesser power levels it could be that this would greatly favor the Me-109G. In wartime accounts, turning contests between P-51s and 109Gs can go on for fifteen minutes in the SAME turn to the SAME side... This is unheard of with Razorback P-47Ds, even with pre-Jan '44 needle blades: except to the right, the Razorback always quickly gains the upper hand over the 109G in less than 3X 360s°... The FW-190A, as should be now obvious, is a whole other matter, and was widely accepted by everyone at the time as the better turner, but inferior to the 109G in high speed handling, especially dive pull-outs (see Russian descriptions of the 190A's dive pull-out...).
I think it was Gunther Rall who compared them as such: "The 190 was a broadsword, the 109 a rapier". A Broadsword is traditionally seen as being swung in a curve, while the rapier is used in a forward straight motion... A very apt comparison...
Once prejudices are ignored, the overall picture becomes quite clear.
Gaston