The question was whether it was a clinically relevant metric - it is absolutely a useful one. You are correct that it is not an indicator of percent of body fat, it was not designed to measure this and using it for this purpose is mislead. But there's a world of difference between "it's bad at measuring body fat" and "BMI is pseudoscience". It's unfair to characterize it as lacking scientific rigor because there are plenty of scientifically rigorous studies involving BMI. It is extremely useful as a clinical indicator of one's health, in the same way that body temperature can tell us things in the context of other metrics and can also tell us some high level information about a person's general health.

But perhaps most importantly, it's extremely useful when we come to population health where generalized indicators are often more useful than hyper-specific ones. Indicators which are easy to measure and gather from relevant data sources are also often more useful than ones which may be more accurate on a per-individual basis, but less important when measuring the health of entire populations. I apologize for any condescension in my comment, I was suggesting that you become more educated in matters of public health because indicators like BMI are invaluable in this space.