Only problem with that is BMI is a terrible gauge for obesity, almost to the point of being medically useless. It presumes everyone has exactly the same build, muscle mass, and fat distribution. A better gauge is the ratio of fatty tissue to lean muscle mass, something that must be calculated on an individual basis.
It also has little to do with actual health. A person with a "perfect" BMI (18.5-24.9) is statistically more likely to suffer health issues than an overweight BMI (25-29.9). Surprisingly, persons considered underweight (BMI less than 18.4) suffered the same rate of morbidity as people considered obese (BMI 30-34.9) (source: National Center for Health Statistics, CDC)
That CDC study has been widely criticized for not taking into account all the junkies (e.g. skinny tweekers), alcoholics, smokers, and sick people who have a low to normal BMI because they are sick and can't eat enough. BMI is in fact very good on a
population basis. I saw a study that showed that BMI actually understates the number of fat people because the same value is used for men and women, and by a body fat percentage, people tend to be even more are fat that the BMI indicates. Especially women.
Now on an individual basis, yeah, BMI is terrible. The BMI is not designed for that but used anyway. Almost every man that works a good strength training program for not even a year will get into the BMI overweight category, even at fairly lean body fat % levels.
Regarding the sugar carbs versus "good" carbs, that is insignificant. I would guess that the difference amounts to just a few pounds at most (if anything at all), from personal experience and what I hear of bodybuilders who are cutting. (It is hard to tell for sure because almost all people who diet strictly do it the "good" way.) A few extra pounds makes a big difference to someone in a bodybuilding competition or trying to make their weight class for some other event. However, the problem is that people are typically something like 30 pounds overweight. There are a lot of fat women who claim to always "eat healthy" and yet are overweight, and I believe them. I created a spreadsheet and assuming 3500 Calories per pound of fat and a maintenance level of 4.5 Calories per day per pound of fat (a middle level based on the various estimates I have seen), a surplus of one plain bagel each day (290 Calories) will put on about 18 pounds to an average woman (average heightand a healthy weight) in one year. This would move her from a size 2-4 to about a size 12.
For what to do about it, I think that is easy. Make fat people pay higher health insurance premiums. Since BMI sucks, a simple measuring tape is all that is needed to get a reasonably close body fat percentage.