Well, they most likely ARE cheating. A while back, I got into an argument over nurburgring lap times with ArtM3 (of course, who else???). I put together a regression model demonstrating the extremely high degree of correlation between nurburgring lap times and power to weight ratios. At the end of the day, physics probably has more to do with performance than the color of the badge on the hood of the car.
Except for a couple of porsches that performed real well and the M3 CSL, pretty much every car is within one standard deviation of the regression line. Even those slightly faster cars barely broke out.
Now look at this. As they say, a picture is worth a thousands words.
The straight math would say the GTR's time is roughly 22 seconds faster than it should be. Well guess what, when porsche tested a stock GTR (without extra boost and slicks), it was 25 seconds slower than Nissan's time.