Related: https://forum.effectivealtruism.org/posts/omoZDu8ScNbot6kXS/beware-surprising-and-suspicious-convergence
it's unlikely that veganism is the best diet for animal welfare and for the environment and for individual health and for taste. Mainly because when you're optimising for one thing (reducing harm to animals) it's unlikely you would also find the best way to eat healthy, or have the best taste.
In other words, from the set of all possible diets, searching for the global optimum for one factor, is unlikely to lead to the global optimum for all other factors.
So for veganism, it is conceivable that is has lower carbon footprint than regular diets, but it would be surprising if it had the lowest footprint of all possible diets.
But it could be the case that we care about all these and that on aggregate, veganism is the best diet.
I.e. if our utility function looks like this F(animal welfare + environment + individual health + taste) assuming we care about these equally (we don't) I could imagine veganism as one of, or even the, best.
Made up numbers below
Diet X: 10A + 10E + 10H 10T = 40
Veganism: 8A + 7E + 7H + 5T = 27
Regular diet: 2A + 4E + 5H + 9T = 20
Healthiest diet: 4A + 4E + 10H + 6T = 24
Where A= animal welfare, E= Environment, H= Health, and T= Taste.
I don't think animal advocates usually believe in this, but in the naive version of "Veganism is the best diet on all factors." And that makes me wary of other claims they make like how it's better because plant-based diets reduce antibiotic resistance and ditto pandemics. Or that going vegan may be better in the long-term because of moral circle expansion.
However, this should make us cautious, not dismissive. A small update against self-serving claims is justified, but complete dismissal risks ignoring factors that make animal advocacy higher on the importance scale.
