Unless you are an expert in the field, I'm not sure why I should believe you. Of course companies are going to publish only results that look good. But a lot of scientists are going the pre-registered route, and claiming that hundreds of studies are wrong seems like it would require extrordinary evidence.

I just talked to my wife about this, she was doing some research tdcs research as part of her PhD 3-4 years ago. She say's that extremely precise electrode placement is necessary for P < 0.05 effects, so that might support your position that retail tdcs devices are largely bullshit. But the research isn't about commercial devices (largely), it's on lab subjects who have properly placed electrodes.

Admittedly, I'm not an expert in tdcs. My experience is limited to what might be considered the typical, prevailing, suite of neuro-electrophysiology methods: stimulating electrodes + glass pipettes into acutely sliced rodent hippocampus/cortex; same for organotypic slice cultures; in vivo implantation of multiunit recording electrodes + fiberoptics into rat cortex. I think these methods provide enough background in ephys to formulate an educated opinion on tdcs. Also note that I'm just as cynical, if not more, when it comes to research reports using those established methods I mentioned above. Suffice to say, it's not the methods that concern me, it's reports of some marvelous thing that happened while using such methods.

You are right though, the claims about the utility of these commercial devices (and really any non-implanted device) that are by far the most dubious.