They found a 15 point MADRS change in the placebo group, that's huge, it's only a 60 point scale, and more than the average SSRI produces. Either the procedure itself is doing something or something isn't right with the study.

Also one issue with all of these studies is they only look at averages and don't do subgroup analysis. It may be that a few patients have an underlying condition causing depression that is highly responsive to these interventions, while it has no effect on the others.