Calories-in vs. Calories-out is not helpful advice. A person's metabolism can slow to a crawl if the person is starving.
It's much better to fix the metabolism, and remove metabolism-destroying foods from the diet. Many common foods should not be eaten (tofu), or only eaten when well-cooked (kale) [1].
[1] https://en.wikipedia.org/wiki/Goitrogen#Goitrogenic_foods
"You don't need to exercise, you don't even need to eat healthy. Eat less than your BMR and you WILL lose weight."
Your claims of a metabolism slowing down "to a crawl" if a person is hungry isn't helpful in the slightest; comments like this demoralise people. They work out, eat kale, make sure to never feel hungry lest they "slow down their metabolism to a crawl" and still aren't losing weight. "why bother?"
Personally I have been through 6 different kinds of diet, calorie counting or not. The only thing to put in a dent long term was a very low carb high protein ketogenic diet (but it broke liver enzymes and cholesterol long term, took it 2 years. Exactly at the point where I achieved the right weight too. Yes, it was nutritionally verified.) and a hardcore CR diet which was terrible to adhere to. (essentially hungry all time after two weeks, preoccupied with food enough to be distracting at work. Also nutritionally verified.)
And in our culture you get to tell everyone you will not partake in their food which is extra torture on top.
So yes, the advice is garbage, does not work in general long term in my long and varied personal experience.
Also, obesity is a syndrome. There are many causes and diet is probably very ineffective for most. At best a symptomatic treatment.
To be at that level of starvation would require the person's body fat percentage to drop to essential levels. "Starvation mode" in normal weight and overweight people is a pernicious myth.
Metabolism control seems to be based of medium term intake rather than fat stores. Either that or there is a set point which is quite hard to move. Or both. Handling the fat does increase basal metabolic rate to an extent. But only so far. The abovementioned slowdown starts with mild thyroid suppression and shows up in bloodwork. This can cut BMR by up to half without starvation. (25% calorie cut from high levels is nowhere near enough)
The problem is, this reduces will to both diet and exercise and induces a low grade depression.
I only have my own metabolic panels taken at my expense weekly during a vegan diet (vetted by a dietician) attempt. A failed one to boot. From correct baseline to borderline hypothyroidism in a matter of weeks. Quick rebound when the attempt was stopped.
The study would have to be careful to exclude people with known preexisting conditions that cause obesity and only take in "idiopathic" subjects.
The most interesting related study I found is the Right Track long-term study of adolescent health related behaviour and total factors. It is currently in progress. Hopefully it won't disappear.