We can assume that the tester, the VTT Technical Research Centre of Finland, is credible for now.
Of course they are credible, but all they do is to perform the tests as specified by their client, and supply the report to the client.
It is then up to their client (Donut Lab in this case) how they use the report. Now if Donut Lab chooses to publish the report as a whole, then we can of course see what was tested, and how it performed. But classically in this sort of scam the idea is that the scammer does not publish the report as a whole; the convenient "minimum" is to simply say that "our product is tested by VTT". Another variation of that scam is to openly publish the report, but rely on the fact that most of the target audience are not technical enough to understand the significance of the test, and rely on the mainstream media bubble. I mean, if the report says that a cell weighing 105.3 grams had integral discharge energy of 22.8Wh, any engineer can calculate that as such their central claim of 400Wh/kg was bogus, but yellow media journalists are not capable or willing to do that, but continue hyping the story, because the story creates clicks for them.
Or, they can simply ask VTT to measure the capacity and temperature rise of the cell, and not ask them to weigh it. Then the most central claim about energy density cannot be assessed at all, and people who question that fact are just stupid and jealous.
But yeah, one thing using VTT rules out is the type of scam where secret wires are connected under the table, invisible to the camera. Or photo editing false numbers on video, etc. And this is exactly the reason to use such testing lab instead of doing your own youtube videos: this is useful for both serious players, and scammers.