Nutrition studies are often wrong and biased, so they can't settle this
Studies can be flawed, and some have been badly wrong. The fix is to compare their methods and see where independent lines of evidence agree, not to set studies aside. What replaces them, usually personal experience with clients, has every bias that studies are designed to control.
Nutrition science has a real track record of confident advice later reversed, on dietary fat, eggs and cholesterol among others. Most of it is observational, riddled with confounders like processed food and lifestyle, and sometimes funded by interested parties. Given that, it is reasonable to hold the studies loosely and give weight to what actually happens to the people in front of you when they change how they eat.
It is true that individual studies get things wrong. Nutrition research leans heavily on observational data, confounding is hard to remove, and funding can shape what gets asked and published. None of that is in dispute.
But the conclusion does not follow. When studies are fallible, you judge each one by how it was designed and check whether different kinds of evidence point the same way. Long-term cohorts, substitution analyses, controlled feeding trials and mechanistic work come at the question from different angles with different weaknesses. When they agree, the agreement is harder to dismiss than any single paper. That is how the link between smoking and lung cancer was established, almost entirely from observational data, against industry claims that the studies were biased.
The alternative on offer is worse by the same standard. A practitioner seeing clients improve has no control group. There is selection in who comes and who stays. Follow-up is short. Results are confounded by weight loss and by cutting out processed food. Expectation plays a part too. These are exactly the biases good studies try to remove. 'Thousands of people' counts as anecdote at scale, not as stronger evidence.
The argument is also applied unevenly. Studies are called unreliable, and then a review of studies is cited in support. If the review is to count, the method of weighing evidence has been accepted after all. The question becomes which studies are better, and that has to be argued case by case. Reviews that rate the evidence on red meat as low-certainty are saying the size of the effect is uncertain. They are not showing that meat is protective.
And none of this touches the ethical case. That rests on whether harming animals is necessary. It needs a vegan diet to be adequate, not superior, and the major dietetic bodies already say it is.
Treating a personal story or an isolated case as though it outweighed systematic evidence. Individual cases are real, but they do not establish general patterns.
How it applies here: The speaker set aside population studies because they did not match what he sees in his own clients, treating uncontrolled personal observation as outweighing systematic long-term evidence.
Applying a principle generally while claiming an exemption for one's own case, without giving a relevant reason for the exception.
How it applies here: Having said studies in general are unreliable and biased, the speaker then cited a review of studies as support, exempting the evidence he liked from the doubt he applied to the rest without giving a reason.