Do different foods have difference metabolic properties, including variations in accessible calories? Yes.
Can we accurately measure calorie intake? Kind of. Even with a food scale, I'm not sure we get to within 10% of the "correct" calorie count because the methods for determining the calorie / lb may have issues.
Could people do weird things, like wrap food in cellophane, burnt it before eating, or even just throw it all up? Of course.
But so what? We have a real world problem here in that Americans (and others) are suffering a material degradation in quality of life due to health issues related to diet and exercise.
There are tons of optimizations possible, but they don't matter to cohort in question. For them, CICO is good enough by far. These optimizations are like trying to teach matrix math to someone who can't add--and doesn't want to learn.
Yet, I regularly see these discussions veer into the realm of these optimizations, presumably because the real problems are too hard and/or depressing.