That article is slightly misrepresenting the actual notice from the Commissioner for Consumer Protection:
formatting link
That document doesn't say anything about "proof" that the devices don't work, just that those companies making them hadn't a basis for claiming they did. In other words, presumably, those companies hadn't done satisfactory testing themselves. It's the same as in Canada where two manufacturers responded to the same thing with actual tests proving that their devices did work, and their authorities relented.
That's explained very clearly in this letter sent by the Canadian Competition Bureau:
formatting link
So I intend to copy the design of those Canadian devices.
EXACTLY, that's why I'm interested in doing my own tests to prove it to my own satisfaction. Frankly I'm still skeptical of the Canadian tests, but short of finding other real experiments documented (and I only found those Canadian ones by clicking through two other Wikipedia pages about rust in general), they're the last word on the matter. That I can find.
Actual tests:
formatting link