Maybe....
They could use the "per capita" type of statistic to fine tune it down to the gnat's ***...
Don't look at the percentages of EV versus gas fires, look at the amount of EVs on fire versus the amount sold and owned, then compare that to the same figures with gas and diesel cars.
If 1 in 100 EVs catch fire and 1 in 1000 gas/diesel cars burn, THERE is your evidence.