I never said it wasn't used. I said it's not accurate. Although the athlete argument is overused, it's true. My given example uses actors. Tom Cruise is not and has never been obese during his time on screen so how do you explain that one? BMI uses a fixed formula and does not take different variables into account like muscles and adipose tissue in relation to the rest of someone's body.
To be honest, people can believe what they want, but when the creator himself says it's not accurate and one shouldn't rely on BMI alone then I think we should listen to him 
As I also said, BMI is an indicator. Nobody's wrong for saying "people with a BMI of over 30 are more likely to have problems with XYZ during pregnancy" or "people with a BMI of over 30 are often found to have problems with XYZ" or anything else that points to the fact that that category is often faced with certain problems, but it shouldn't be used alone because it's not accurate enough. The fact that the NHS has it on forms doesn't mean it's any more credible. There are, and will always be, people who believe it works. There are, and will always be, government bodies and institutes that use it, and that's fine. But the NHS don't rely on BMI alone to tell someone whether they're obese or not.
BMI is not accurate enough to lean on solely as evidence for obesity which is why I thing the previously mentioned argument is invalid because it's not drawing on proper medical evidence, but a formula thought of by a statistician.
:)