There's a sunk cost already spent for an ICE car that's already been produced. There's an opportunity cost to swapping to an EV immediately. My point is simply that the situations are complicated enough that the only reasonable "one size" approach for a heuristic to balance those costs is one along the lines of "replace your ICE car when it's reached the end of its useful life, and replace it with an EV".
No, this probably won't be the best overall. That requires individualization. Someone still clinging to a 40 year old gas guzzling truck would be better off scrapping it. Someone who bought a sedan in, like, 2017, it still has a few years of well performing life in it would do best to keep it til it dies and then replace with an EV.