I guess high quality is relative. Of the 10 underlying studies, at least one of them has been retracted (also happens to be one that the authors rated most highly on quality), majority were open label, and there are indicators of publication bias in the articles that were included. The point estimates for weight loss (~15lbs) given the duration of the underlying studies is improbable, so of course deserves some scrutiny.
I didn't dig into all the studies, but here is the retraction that popped with literally the second underlying article that I looked at. Not a good look when the only underlying study that was published in a journal family you would recognize is retracted.
https://nutrition.bmj.com/content/early/2025/09/23/bmjnph-2023-000823ret
Wide ranging concerns were raised about the quality of the work after publication, including the approach to statistical analysis of the dataset; implausible values; concern about the reliability of the underlying dataset; inadequate reporting of the work; and absence of prospective trial registration, which is a breach of BMJ policy.2 3