So don't call it race. Mother Nature likes to create semi-hemi-demi-species and telling us that our neat Linnaean classification system has fuck-all to do with the real world.
Still, if you are trying to get clues after a crime, and there is blood spattered on the wall, that blood can be DNA tested to tell us a few things about whomever that blood belongs to. Not just sex, but also what part of the world that person's ancestors came from. So don't call it race, that doesn't eliminate what little genetic diversity the human species does have.
The problem with "race" is that it obscures, rather than clarifies, biological descent. We have known how reproduction works scientifically since the 1930's, and folk classifications of "race" have been scientifically meaningless ever since.
Your forensic scientist is not, in fact, useful if all they can give you is a "race"; the vast majority of people have a much more complicated genetic pattern than belonging exclusively to one apparent race would suggest. Real genetic science is both more precise, and more ready to account for phenotypic variability. Your blood does not, in fact, tell you "what part of the world" you came from, but rather what reproductive networks your ancestors belonged to, as a result of social and political patterns that may have originated in geographical proximity at some point, but were in all likelihood quite complicated in the last few (and therefore most important to creating positive id) generations.
Is it more useful to know that someone "looks black", or that they are one of only a couple dozen people in town, all related to one another, with lineage tying them to a particular village in Ghana, regardless of what their expressed skin color happens to be? The whole idiotic notion of race gets in the way of correct descriptions of inheritance, and new forensic anthropologists usually have to be trained out of making race-based assumptions before they become useful lab techs.
This doesn't make race unimportant. As a social and political reality, it is both empirically real and important for a social scientist to understand. It just doesn't have anything to do with biological science.
