The problem though is that serving sizes typically don't reflect how much you actually eat, especially for food that's bad for you. A serving of chips is usually a fistful, a serving of chocolate is not the entire bar, etc. Overall the system is ok, but there's room for improvement.
On the other hand, nutrition labels are such a small problem, the big problem is that people eat too fucking much, and no label in the world are going to stop them from doing so.
http://www.examiner.com/x-672-Disease-Prevention-Examiner~y2...
A lot of food says 0% fat, when the ultimate result of all that sugar is essentially fat.
The problem is that people don't know how to interpret the labels. Better education would help with that. Weight watchers, for example, has their points formula which boils down the calories, fat, and fiber values into a single number that is easier to understand. Some food packages even include the WW points number -- but it wouldn't be very accurate or scientific to mandate that sort of thing.
From many kinds of studies conducted over years, we are quite confident now that a calorie from fat will cause a similar amount of weight gain as a calorie from carbohydrate. There are some interesting questions about whether eating carbohydrate calories versus fat calories will make you eat more calories, but based on what you put into your mouth, it's pretty clear that the source of the calories is really not important.
http://www.time.com/time/health/article/0,8599,1822118,00.ht...
The food that contains 0% fat doesn't contain any fat. That's actually useful information. If people think that 0% fat foods are automatically good for them, that's an education problem not a labeling problem.
As cwan mentioned earlier, read Taubes.