Interesting. I'd be interested to see what would happen to those numbers if teenagers and people over 65 were removed from the equation. Interesting to think about ramifications that don't come across in the numbers, like how income factors into people's decisions about what vehicles to get, how personalities alone determine which type of vehicle to get---like, a 43-year-old actuary who is pragmatic and a safe driver isn't going to have a Nissan 350Z as his first choice of vehicle. And look at how, in many instances, foreign cars almost ALWAYS do better than domestics. Does this indicate an inherently more safe vehicle, or are people who purchase foreign vehicles just more likely to be more cautious drivers? I wonder.
In a way, these numbers actually mean very little on face value. Buying one car or the other isn't necessarily going to improve/decrease your chances of being in a fatal accident, and people might read too much into these numbers. You can do almost anything you want with numbers, make just about anything look better/worse. But it IS very interesting. More information (breakdown of age/sex/income/education, for instance) might yield more usable information. Just my opinion, I'm not a statistical analyst.