Alex:

No, I don't...that would be interesting information indeed.

Tom Passin:

I've discussed those very subjects in the previous installments of this series.  Note: the adjustments you're referring to are "Atwater factors", which were discussed in the comments of Part I. 

And yes, you've grasped the point: once you multiply the intrinsic inaccuracy of "calories" by the known errors in measurement, the error term in the numbers you come up with is much larger than the changes you're attempting to track!

Ash:

That's interesting...the Harcombe article points out that "3500 calories per pound of body fat" is an approximation subject to great variation.  I wish she hadn't stopped there, though, as she's implying that the rest of "calorie math" is still valid!

I used to eat bunless McDoubles on the road...now I usually just plan ahead and bring food in a cooler.  Or, if I'm really caught short, I'll stop at a supermarket and make some steak tartare from meat, egg yolks, salt, and pepper.

Bill Lagakos:

That's actually why I don't support labeling laws: not only do they encourage minimizing the wrong things (e.g. saturated fat, cholesterol), they encourage restaurants to give me less food for my money. 

I've always thought "diet" TV dinners were a great scam: they turn tiny portions into a selling point.  "Spend the same amount of money and receive less food!"

Arturo:

Thank you for the vote of confidence!  Keep spreading the word.

More soon!

JS

Sofia:

First, I only make tartare with hamburger that comes from "shop trim", i.e. that has been ground from trimmings cut on site, unless it's a huge emergency. Some supermarkets label their shop trim, but with some you have to ask. If it's not labeled, I assume it's from a big plastic tube, reground to look fresh (which is where most of the "fresh" ground beef in the case comes from).

Keep in mind that no one in the USA has ever died from eating meat from the meat counter in a grocery store!  The majority of deaths come from contaminated fruits and vegetables (cantaloupe (53), fenugreek sprouts (30), spinach (3), green onions(3)), with one notable instance of dodgy Mexican cheese (~50).  The only deaths due to meat have been from industrially-packaged cold cuts (~40) or Jack in the Box (4), with one single exception: a 5-year-old child in Wales.  (Source.)  And I note that the E.coli outbreak in question was in school lunches, which are very well cooked.

That being said, I don't recommend it to anyone else, particularly anyone pregnant, very young, very old, or otherwise with a compromised immune system.

Jen W:

I believe you're correct: nutrition labels are generally for the food in its raw state.  For instance, bacon supposedly has ~90 calories per slice, 80 of fat...but a substantial amount of the fat is lost as it cooks.

JS

edster:

Exactly.  CICO zealots often use that particular strawman: "You don't believe in BASIC PHYSICS." 

Of course I do! 

The difference is that I understand that the human body is not a heat engine that runs at a fixed rate and at fixed efficiency regardless of fuel source, genetic and epigenetic history, and environmental conditions. 

Furthermore, I understand that the combination of labeling and estimation errors cannot possibly produce a result accurate enough to produce the results predicted by the naive CICO model.

JS