A disputed study raises a practical question. If population maps put people in the wrong places, could disaster plans underestimate how many need help?