I know nothing about the physics of it but would that not depend on the causes of fires? I may be misunderstanding but it sounds like you're implying that fires are more likely in older cars - maybe that is actually the case (I'd be willing to believe it).
A huge number of vehicle fires are electrical in nature.
I've been a firefighter for ~10 years. Nearly every vehicle fire I've seen falls into one of three categories:
1) Deliberately set
2) Faulty/old/corroded wiring (by far the most common)
3) Some flammable material coming into contact with something really hot (commonly after an accident where a car has gone off the road into tall dry grass)
Fires due to the fuel tank failing are _extremely_ rare (I've never seen one, or even heard of one in neighboring areas).
It may be related to the fact that I'm in an area that uses lots of nasty stuff to deice roads, but corrosion is a major issue with vehicle electric system, and the older a vehicle is, the more likely it is to experience a 'spontaneous' fire.
Another important effect is that some cars have known defects putting them at increased risk of fire, and some of those cars are still on the road, which inflates the statistics for the average automobile.
