There are criticisms of it, some of which are in the criticisms section of the Wikipedia article, and look to be valid.
I never liked when people talked about debunking an experiment that was honestly done and just needed refutation or at least correction. If researchers publish research that was honestly done and then later found to have statistical flaws, or that is disproved with a larger sample size or whatever, I wouldn’t say it had been debunked. I’d say it was refuted or corrected or refined or whatever. If it was a “research” study where the group publishing it knew is was junk designed to promote an agenda or sell a product, then the term debunked is fair.
It's all fair. All studies are flawed but some are useful. As a paradigm shift for addiction I could see it being less than a home run. As a statement about happiness it's all well and good if for no other reason than an analogy and potential positive finding.