Completely inane article. The system of measuring kcal is done as a measure of pure energy and that's why it's still used. Specifically, it reflects the number of kcal in the given macronutrients which are then added for total caloric count. The reason macronutrients are not and should not be adjusted is that the thermogenic effect of food (what the article refers to as 'cost of digestion' -- which is a misnomer) is an increase in metabolic rate and not a reduction in the energy provided by the food. Protein doesn't always provide 4kcal/g just as it doesn't always provide 3.2kcal/g. They're both ballpark estimates and individual foods provide different amounts of energy for different people. It's much easier to measure calories in food by burning them as in the current system than measuring them in vivo due to differences between individuals. Aside from that, there's minor technical details which are assumed to always be true which are not (e.g. urea isn't always produced by protein consumption, etc.)
