This link seems highly suspect. In one graph they seem to be comparing number of vehicles recalled due to a potential fire hazard. That is not the same as actua

This link seems highly suspect. In one graph they seem to be comparing number of vehicles recalled due to a potential fire hazard. That is not the same as actual number of vehicles that caught fire. You might have a model with 500,000 vehicles recalled because a faulty switch caused a fire in 5 cars. It also tries to combine data with hybrid vehicles with ICE vehicles because they also have a gas component. They don't do the same with EV, even though hybrids also have an battery component. There is no attempt to distinguish if the fires were caused by the gas or the battery in those cases. When I see a study fudge the numbers in one area, I am suspect of all the numbers it presents.