The P-51D performance has been discussed ad nauseum.
At 50 percent fuel the 3US version is a dream.

The 3US 47D-26 is a major performance problem.

In a confidential report, 29 May 1944, Maj Gen. Kempner, VIII Fighter Command, Deep Fighter Escort, Merriam Press, Military Reference Library L-45, stated that the 47 had broken the back of the Luftwaffle and that the 51 would finish the job.

The 3US version is a DOG so what's the problem?
The problem is that the underlying GMAX model doesn't do justice to the 47.

My settings:
[flight_tuning]
cruise_lift_scalar=1
parasite_drag_scalar=1
induced_drag_scalar=.6
elevator_effectiveness=1
aileron_effectiveness=0.95
rudder_effectiveness=1
pitch_stability=1.05
roll_stability=1
yaw_stability=1
elevator_trim_effectiveness=1
aileron_trim_effectiveness=1
rudder_trim_effectiveness=1

USAAF performance data 47D is 314 mph IAS Mil power 10k.  Original 3US is close enough.
USAAF performance data climb Mil 10k is 305 IAS, 2,400 fpm and at 20k 246 IAS and 2,100 fpm.
Phenominal climb for a WWII bird 3,000 lbs heavier than the F6F.

The eliptical wing and well designed turbocharger gave it extraordinary performance above 20k.  Much better than N fighter performance.

The 3US model doesn't come close out of the box, given poor dogfight performance.

The 47D-26 was a much better fighter than represented in the sim. 

See also www. flight journal.com
P-47 Bolt Special Issue esp. for Corky Meyers assessment.