They do serve as a good indicator. The primary problems that occur during evacuations are "people don't believe it will hit them" and "there's no way to move all those people." I'd also doubt that our math is good enough to precisely calculate the impact point of a meteorite more than few days in advance. Hit/no-hit probably, and which region, yes, but off by 50km? Meteorites break up on atmospheric impact, getting the precise point right is probably impossible.