BMI is useful for historical population comparisons because you can calculate it using just height and weight and it's already been in use for a very long time. It's so crude as to be very misleading when applied to individuals, especially if you decide to turn your brain off when deciding how to evaluate the information.

The origins of the calculation are immaterial. It's value is in comparative studies, not direct judgement. The actual judgement of "good" vs "bad" BMI numbers is dumb(ish) but it is good for comparing populations across both time and space.