This is generally incorrect, because if you think about it the cars that died young aren't included in the average. The problem is that there are two averages being cited, and you can't just multiply them to determine a joint average without knowing the nature of the underlying distribution.

Multiplying averages doesn't work for general distributions. (It probably works out mostly alright here.)

Not only is it a math error to multiply it out like that, but it's also wrong to extend the logic to electric cars. Electric cars are likely to have LESS than the AVERAGE CAR total mileage, because (a) they're less useful in places where people need to drive extra-long distances, (b) the batteries get worse and worse at holding a charge over time, just as cellphone batteries do. (c) it's still a rapidly-advancing technology, meaning older models will get obsolete quickly and need to be retired faster than they do with a more mature technology.