BMI is a good tool for statistical analysis of populations, it is not particularly useful for individuals. Mr C. once managed to get well into the "overweight" category with his body fat in single figures by weight training, just for the heck of it, and ran several Ironman triathlons at that weight. (He got smaller when he did the double Ironman because it really makes a difference to the joints when you're running 2 marathons back to back).
The interesting thing is that if you read the scientific papers rather than the official recommendations, the biggest factor in health is not weight or fat but being active. The deaths from all causes tables also show that the obsession with being thin is not actually supported by the science, the risk of ill health doesn't start to rise dramatically until well up into what is usually listed as obese and the lower end of the normally recommended weight ranges is actually at the point where overall health risks are starting to rise, albeit only slightly. Given reasonable levels of activity the range of weights and sizes that are basically healthy is much broader than is sometimes suggested although clearly there is a point at which being very fat really is a serious problem. Bear in mind too that the usual propaganda only looks at heart disease and diabetes and ignores a wide range of other problems and in some cases being overweight leads to longer survival and that includes some counter-intuitive situations. (Read the paper linked below for some interesting details).
My suspicion with respect to the official guidelines is that they are looking at it from a purely financial point of view in terms of cost to the NHS, rather than numbers of deaths. Diabetics and people with heart disease who live for several decades with their problems cost the nation money. There is therefore incentive for them to make recommendations that will reduce the cost, not necessarily early deaths. Cynical? Yes. And I've had far too much to do with the finance industry and political affairs to dismiss that out of hand.
Although being overweight is frequently touted as being the cause of diabetes, heart disease etc, the causality is still not clear and there is some evidence to suggest that in the case of diabetes in particular, early stage type 2 diabetes is the cause of fat retention rather than the other way around.
Daily intake recommendations are also effectively meaningless given the huge variation between individuals. In my 20s when I was an international level dinghy sailor I was doing huge amounts of exercise, both strength and aerobic yet my caloric intake to maintain my weight at a steady level was well under 2000 calories per day. Now that I am old and less active that has dropped to around 1400 per day and yes I really have kept detailed food diaries.
Having said that the recommendation to stop eating sweet baked goods and junk food is good simply because they give huge amounts of calories in a form that tempts you to eat, and eat, and eat. (You don't need to guess how I know). If I want to lose weight the least painful method is to go low carbohydrate and after the first 3 or 4 days of mad cravings for sugar it settles to a relatively tolerable level of hunger.
As far as general dietary advice is concerned I like Michael Pollan's recommendation to "Eat food, not too much, mostly vegetables."
An interesting link.
http://jama.jamanetwork.com/article.aspx?articleid=200731