Earlier quoted context omitted.
Yeah, BMI should be retired as a useful metric, percentage body fat is far more useful. The problem though is measuring this accurately... (the scales that include the body fat percentage give wildly different numbers depending on time of day, water consumption etc.)
It is a useful metric, but it was never intended to be used for assessing individuals. It's intended use is as a metric for populations of people. It works very well for that.
It's pretty useful for individuals so long as you make adjustments for people who exercise every day (ie, very few people) or people over about 6'5" (again, not many people) and very obese people (where it tends to under-estimate obesity).