He's saying that this is a well established fact regarding every fad diet. Usually in the sciences, a new method measures itself against a "control", which might be a placebo or non-treatment, but might be another well established treatment, such as a drug for the same disease. For a fad diet, a sensible "control", or "null hypothesis" would be another fad diet, not involving potatoes. One would only have grounds to say that this fad diet worked if it worked better than the best alternative fad diet.
I'd argue it's a well established fact of any diet, regardless of how it is lost. Changing your body's natural weight long term is very difficult no matter what you eat.
