So my hope and assumption is that they were likely saying "for individual use it's not great" which of course is PCP's bread and butter and on that note I would agree (with some caveats).

Unfortunately I'd like not to doxx myself by showing my own published stuff and having my name out there, but I actually got frustrated at one point in grad school by looking for something novel to say about the predictive value of WHR, circumference, and BMI and after a lot of hours on a really big longitudinal study I came away with "meh, use whichever".

Of course, if I wanna be super accurate, I'm doing a dexa scan or something, but geez human research is already expensive and I could rent this multi thousand dollar machine or just have a cheapo scale and wall height measurement and have it be good enough for research purposes.

Lastly, I think it absolutely must be said that regardless of size, moving more can counteract a ton of this stuff to the point where the negative cardiometabolic effects are hugely diminished. A lower bmi person is absolutely not 1:1 healthier.