Maybe posted before, but here is Corky Meyer's writing in Flight Journal:
"To simplify the evaluation and reduce data, we decided to test-fly the Hellcat and the Corsair in close formation. Instead of comparing complex calculations, performance could then be compared directly at the critical altitudes of the main stage, high and low blower altitudes of the engine's superchargers, and from cruise to high-speed, level flight with water injection. We also included some formation dives to learn which airplane was the slickest.
Except for the Corsair being 20 knots faster than the Hellcat in the main, sea-level, supercharger stage, both fighters had almost exactly the same speed at the low and high blower stages from 5,000 feet altitude up to service ceiling! In essence, they had the same performance. Our formation flights showed that both airplanes (with similar power settings) were in closely stabilized formation at all altitudes tested above 5,000 feet Sometimes, the Corsair would slowly gain a lead of 100 to 200 feet after five minutes of stabilized power flight, and sometimes, the Hellcat would do the same. Considering that both airplanes had the same engine, propeller, gross weight, wingspan, etc., they should have had about the same performance. We did notice that during these runs, the Corsair always had about a 20-knot indicated airspeed (IAS) advantage! We didn't realize just how embarrassing it would be to solve that dilemma.
The reason the Corsair was faster in the main stage blower was that its engine and carburetor were provided with ram air coming in directly from the forward-facing wing duct, whereas the Hellcat had the carburetor air coming in from the accessory compartment of the fuselage just behind the engine, with no ram air effect Our airplane was getting carburetor air at the same pressure as it would have were it motionless on the ground, and the Corsair was getting carburetor air supercharged by the speed of the airplane giving it more power (speed) in the main stage blower. In both aircraft, however, the designs were similar in that they provided ram air to the low and high blower stages. Our engineering department defended its position because taking the warmer air for the main stage blower would prevent inadvertent carburetor icing engine failures. Many Wildcats that had ram air in the main stage like the Corsair were lost because pilots failed to take precautions in time to avert this type of disaster. The Hellcat design was reviewed and approved by the Navy. I had had a carburetor icing accident during final approach on my first flight in a Wildcat a few months previously; it resulted in my first deadstick landing and a vertical ground loop. I therefore heartily agreed with the Navy's decision.
After noting the 20 knots indicated airspeed difference that had caused all the "lower performance" ruckus for our Hellcat, we eagerly decided to change the airspeed system so that it would read evenly with the Corsair when they were in formation. We had taken a lot of flak from all who had flown both airplanes (but not in formation) and, therefore, everybody 'mew' that the Hellcat was inferior in high-speed performance. We liked our simple and less complicated airspeed system with the static and dynamic orifices on the same boom, but we decided to go whole hog and put the static orifice on the fuselage (like the Corsair) to tailor the system to read 20 knots higher. We tried several orifice locations to get the required reading. After I had done a thorough testing of the final system over the entire flight envelope-or so I thought-I proudly flew the airplane to the Naval Air Test Center at Patuxent, Maryland for an evaluation. We soon found out that we had not purloined the Corsair airspeed system design thoroughly enough.
We soon received the Navy's glowing report of the new system; and it went on to say that the Air Test Center had never tested an airplane with such remarkable low-speed performance in its entire history. They found that in a left side slip with the wheels and flaps extended, the Hellcat could fly at zero airspeed. Wonder of wonders! Grumman led the industry again! Upon re-evaluation, we found that the engineers, inexperienced with flush static airspeed systems, had designed ours with only one orifice on the left side of the airplane, and it was very unbalanced with the flaps down. As the senior engineering test pilot, I was in deep doo-doo for not testing the new system in all side-slip conditions. A dualorifice system way behind the lowered flaps (similar to the Corsair's) finally provided a satisfactory means to give the Hellcat a cockpit indicated airspeed reading comparable to the vaunted Corsair's. That was the last we heard of the Hellcat's performance gap with the Corsair. Performance case closed. "
