Given that we know that diets and changing habits doesn't have lasting effects, what doesn't sit well with me is to risk my health to avoid taking a drug that helps.
People talk past each other on this because for an individual it technically can work, and sometimes does, but on a population level, as extensively studied by people whose job it is to study that, it definitely does not work. Even with tons of regular attention from professionals and a cohort selected to have above-average motivation, it’s surprisingly poorly-performing (and that’s a crazy expensive level of intervention)
Think about how we describe contraceptives, medically speaking: a failure rate is tracked and promoted that’s the in the wild rate of failure, not the ideal-use rate of failure (which can be effectively zero!). The diet and exercise equivalent of a contraceptive couldn’t be sold, because its failure rate would be way higher than its success rate.
So “we” (people who’ve paid attention to the science on it) know it doesn’t work (on a population level), like for-sure definitely does not work, but a person reads this assertion of fact and goes “but wait it worked for me, this person must be dumb or something” but that’s not it. It’s two different perspectives on what it means for it to “work”.
I think it'd be more accurate to say that informing people that they should change their behavior doesn't work. A person changing their behavior can escape obesity. The reality is that most people won't be able to change their behavior without some other kind of assistance.
Even with extensive assistance (which is way too expensive to widely deploy) outcomes are weak.
Last I checked, researchers in the specific area of high-touch weight interventions were excited that they finally had a tool that might more-than-barely work… in glp-1 agonists.
However, all the evidence is that the vast majority of people fail at changing their habits in ways that produce lasting weight loss, so it does not generally work as advice for reducing your weight.
So you're technically right, but it is irrelevant, because we don't know how to actually get people to change habits with any meaningful rate of success.
At this point it is downright harmful and wildly unethical to recommend it when we now have a far more successful option.