The Sauber C9 hit 248 mph on a qualifying tune. It did not hit that speed in race trim. The C11 (the successor to the C9 which used the same M119 engine) was rated at 950 hp in racing trim, very far from the "1,500 hp" you're claiming.
1500hp or thereabouts in qualifying trim is a well documented number given by the race team in many 1990's interviews, ill provide you the links. Whether they over embellished or not is not for us to judge as I'm sure neither of us has had anything to do with the cars engineering and development.
The GT-one was a LeMans Prototype (very different from the Group C regulations under which the Sauber C9 was built) and in low-downforce race trim it could clock 235+ mph despite having less hp than the C9 due to superior aerodynamics. Remember, the C9 is originated in 1987, the 1989 models, while tweaked and having the newer M119 engine, were not examples of state-of-the-art aerodynamics by the time they were racing. The GT-one is ten years newer and the amount of aerodynamic knowledge gained in those 10 years was HUGE.
I'll put more research into the GT-one but what I've been reading gives a 230mph top speed.
As far as a 750 hp car hitting 225, that's absolutely possible with the right aero setup. The base GT-R with 480 hp hit 195 mph for Motor Trend. I realise that the power required to go faster increases with the square of speed, but it's not at all inconceivable that 270 more hp than stock could result in an extra 30 mph.
I'm trying to find a calculator for this but what I do know is that a 1000hp viper with a .28 cd and lower frontal area went 225mph in the late 90's. I've found formulas but not an easy calculator. So ill do the math just for you. I'm using 6.8 sq ft for the frontal area. Don't try and talk like I don't know what's taken into account when measuring terminal velocity. Just because I didn't site every variable Eros not mean I'm not aware of them. People have a huge problem with wanting to debate when they don't even have the grounds for it. Since your such a physicist why don't we both run the math and then agree to who is right or wrong.
These "odd inconsistencies" are, in reality, neither odd nor inconsistent.