Since I'm now convinced calories matter, I had the following thought this
weekend.  If we want to count caloires, shouldn't we be thinking in terms
of "effective calories" rather than bomb-calorimeter-derived calories? The
proper factors might be:
Carbohydrates = 1
Protein = .8
Fat = .55

Assuming I got those factors close, let's look at Todd's single-sitting
serving of prime rib, which he said has 1500 calories (900 from fat and
600 from protein).   Apply the math and we are really talking about 890
*effective* calories rather than 1500, because of our body's relative
ability to handle each food subtype.  But with 1500 calories of honey,
which gets its calories entirely from carbohydrates, and the effective
calories are still 1500.

The website that Oliva posted the other day (www.nutritiondata.com) has an
absolutely fascinating pyramid the shows the relative percentages of
calories from fat, protein, and carbohydrates for a particular food.
Problem is,if the above is true, it's skewed toward the fat and protein
side of the pyramid since they don't account for effective calories.