I guess I'm trying to argue that Americans are more voluminous, and trying to approximate volume through mass since body volume by population isn't a commonly available stat. I'd argue that mass is closer correlated to volume than BMI.

Usain Bolt and Michael Phelps have normal BMIs. BMI is very sensitive and not very specific, meaning that it has few false positives and lots of false negatives, the opposite of what you suggest.

In this article[1] "Measuring Adiposity in Patients: The Utility of Body Mass Index (BMI), Percent Body Fat, and Leptin" you can see this dramatically illustrated in Figure 1 [2]. The upper left quadrant (different colors for men and women) is the most populous. It contains the people who are obese by body fat percentage but not BMI. The lower right is the least populated. It contains the people who are not obese by body fat percentage but who are flagged as obese by BMI.

The small fraction of the population that have devoted such attention to the cultivation of muscle over other goals, such as athletic performance, are surely well aware of their body fat levels and are unlikely to be impacted by this edge case in BMI measurement. (Their excess muscle may also not be healthy for their cardiovascular system.) Dismissing the utility of a easily applied and highly sensitive screen like BMI is a mistake, both for individuals and for public health.

[1] https://journals.plos.org/plosone/article?id=10.1371/journal... [2] https://journals.plos.org/plosone/article/figure/image?size=...