With regard to possible NTSC/PAL differences, I just ran some tests with my NTSC/NA game:
I bought a new 4.5 mil Toyota GT-ONE, 0.0 miles, and took it to the Top Speed Challenge. I did not do anything to the car, no oil, no settings changes (this includes specifically leaving the ASM/TCS at the default of 10/10/7). I used options settings of AT transmission, active steering off, steering type professional, power assist on, feedback strength mild, and a DFP. The intention here is consistency and repeatability, not fastest speed.
I made three runs, posting speeds of 236.56, 236.06 and 236.54 mph. I then reset the game to keep the mileage on the car at 0.0.
I then bought a new Oreca Viper, again left all settings stock, and went back to the Test Course. I posted these times: 204.00, 204.03, and 204.01 mph, and reset.
I invite others to try the same thing with the exact same settings; at least one other NTSC and a couple PAL. Perhaps we can pinpoint the difference between NTSC and PAL.
Intrigued by the B-spec aspect of this, I took the GT-ONE and Oreca Viper to the Driving Park and Test Course for some practice sessions in B-spec. I did five laps with each car, all pace 3 with overtake off, and came up with the following lap times:
GT-ONE, 1'39.480, 1'39.327, 1'39.299, 1'39.300, and 1'39.282.
Viper, 1'55.203, 1'54.791, 1'54.796, 1'54.795, 1'54.800.
This was done with a 10000/100/100/100 B-spec driver. I repeated the test with a 0/0/0/0 B-spec driver and (astonishingly) all ten times were identical. Again, I think it would be worthwhile for other NTSC and PAL drivers to see if they get the same results. Note that it's especially important to reset the game, preserving not only the 0 mileage but also to erase any possible acquired ability on the part of the B-spec driver (although the results do seem to indicate that driver ability is not a factor at all on Test Course).