There is an assumption that is easy to make about health advice: that the health of the person giving advice is a good test of the correctness of their theories. This is a terrible mistake.
Here are a couple of reasons why it is a mistake:
It conflates what people advocate with what they do.
This goes in two directions. Someone could advocate something that they don't adhere to, or they could adhere to it but do something else that, known to them or not, changes the outcome.
For example, someone could claim to gain fat on a very low carb diet, but not realise that those high-protein peanuts they are eating have signficant carbohydrate in them. Or, on the other hand, they could claim great loss, but not mention that were also taking a hormone that turned out to have made all the difference.
It doesn't take into account what the person is struggling with.
Someone with a strong genetic propensity to gain fat and who has been obese for a long time is simply never going to have the results from a fat loss intervention that someone with a more moderate genetic risk who has not ever been fat.
A hard gainer is never going to get as bulked up as a natural bodybuilder type, even if they find the perfect protocol.
This latter problem would even seem to suggest that people who struggle or have struggled with health problems are better judges of what works and what doesn't. There is some truth to this: If someone with a health problem manages to improve it, that positive outcome suggests that someone else in the same position might also see improvement following the same plan.
Moreover, sick people, fat people, people who can't meet their goals through mainstream advice are often more likely than others to hit upon fundamental scientific misconceptions that are holding them back, or new therapies that are incredibly effective. And yet, a person who has recovered from severe diabetes, for example, may still have health measurements that don't inspire great confidence in someone who is starting out with less severe issues.
However, this is also a mistake!
If you want to evaluate whether someone's health advice is good, comparing where they started with where they are is a good starting heuristic, but it doesn't take the place of evaluating scientific research. There can always be coincidences, or relevant information you are not aware of.
People who do health research and promote their conclusions are aware that their own health is under scrutiny. They can hold themselves up as positive anecdotes of what is possible, and they should.
Anecdotes, like observational correlations, are useful for generating hypotheses. A randomised controlled trial that refutes the hypothesis must be considered better evidence than an anecdote or even a correlation generated from a large data set. On the other hand, a negative anecdote can single-handedly refute certain kinds of scientific claim.
It bothers me when people criticise the theories of public figures based on their personal successes. It equally bothers me when people applaud those who have lucked out genetically, and who can basically do no wrong: they will look fantastic under a variety of abuses.
Take advice from fat people, if the advice is scientifically tenable.
Ignore advice from thin people if it doesn't stand up to scrutiny.