That there is variance in energy expenditure both within a population and within a person over time doesn't mean that a caloric deficit doesn't work. It just means that using a single scalar value (which is usually a gross estimate) to drive your caloric intake is a poor approach.
The body has means to regulate it's energy expenditure to maintain homeostasis, and in some people it can be a hundreds of kcal difference. But if you're trying to lose body fat on a 10% estimated deficit and fail, the conclusion shouldn't be that a 20% deficit will also fail.
For some people, a 50% "deficit" fails. And the entire concept of "X workout burns Y calories" is completely bunk. Again, there have been multiple studies to this effect.