> a lot of women who might be overweight are healthy because different % of fat are acceptable depending on the structure of the body
This is a tired argument. Most people who have BMI in the obese range do not have one of oft-cited exceptions that make BMI an imperfect measure.
Everyone knows BMI is imperfect at this point, but the number of people who have BMI in the obesity range yet have healthy body composition is very small.
Let it be noted that I have said overweight and not obese, if you are in the obese category you are 100% unhealthy (even the bodybuilders who inject stereoirs in this category are unhealthy).
It's important to note that overweight and obesity are not the same thing. Most people are overweight, and from what I've seen of modern studies, the health risk of being overweight is almost negligible.
But being obese is a higher BMI than overweight, and the bar is actually quite low. Lower than most people think. A lot of people think they're overweight, but they're not, they're obese.
> Everyone knows BMI is imperfect at this point, ...
Indeed.
BMI is the best thing that people can readily calculate with easily available equipment (a tape measure and scales either at home, gym, pharmacy, etc) plus some relatively basic maths or sticking the numbers into a website.
Measuring body fat using calipers is better but hugely error prone. Similar for waist/height ratios. Body fat scales can be wildly inaccurate.
BVI is far superior but very few people have access to the equipment needed to measure that.
So we're kind of stuck with BMI as the best "simple" measure.