I believe the limitations of BMI for *individual* measurements are well-known -- although, perhaps, not so well-publicised as they should be, if the story TTT links to is anything to go by.
On the other hand, taking BMI averages over a community provides a safe, easy-to-use and fairly accurate assessment of whether the society as a whole is (on average) over- or underweight.
As long as it's used sensibly, and in the context of the person's health and activity as a whole, then there's no harm in using BMI as a useful guide -- particularly for anyone at the extremes of the scale -- but no more than that. In that sense, I clearly would disagree with informing children of the results of "National Child Measurement Programme", and it should really be restricted to informing (but not determining) government policy on school meals, school PE lessons, etc etc.