Neat can maybe explain a couple hundred kcal variance in most people, perhaps there are exceptions but 50%? I've never seen that in the literature.

Calories in calories out is just the summation of expenditure and intake, just because the body is complex and there are many interdependent factors doesn't mean it cant be resolved to a vector which determines weight gain/loss. The problem is people google a tdee calculator, get some scalar which is likely wrong, perhaps substantially, make lifestyle changes, and then have an expectation of some result in a specific timeline that isn't realistic, and then eat a bunch of sodium, put on 2 lbs in their "deficit", and think the diet made them fatter! Or they read that -3500kcal == -1lb fat, calculate their calories burned from the machines at the gym, and get frustrated when it doesn't work (I'm guilty!).

Weight loss is actually really hard because it really just requires a sustained effort over a long period of time to achieve anything. You might not see any results for weeks as your body adjusts, you get your diet locked in, etc. And since your weight can vary so much day to day, it's hard to stay motivated. Ozempic kind of bypasses these problems. You know what else works? 20k steps a day and eating on a backpacker budget :P