Given how far computer vision has advanced nowadays, I would be interested to know if anyone has done an analysis of the frequency and delta of over-speeding vs the make of the car as a (not so great) proxy for how rich a person is. That might be an additional data point to consider when formulating these laws.