My apologies, but could anyone help me find a half-remembered blog post about veganism which directly addresses this very intuitive objection?
The thesis that stayed with me is that veganism is not "too good to be true" because most of its features are not convergent, but correlated consequences of a single fact: animals are themselves evolved to optimize a goal which is not "being eaten" (in fact, it's almost exactly "not being eaten"!), and they are therefore suboptimal food-generating technologies. Specifically, animals put much of the energy they consume into action and cognition, instead of growing edible flesh; due to this inefficiency, they cause more emissions and require more freshwater and land than plant food; and they suffer when used as food machines, because they evolved to avoid predation. Consequently, we should not be surprised that we can develop technologies for converting plants into food that outperform animals across these metrics. (Notably, health and taste do not clearly follow from the same fact, and veganism is plausibly merely adequate rather than optimal on these metrics.)
Does anyone recognize the post I'm trying to paraphrase? I can't find it on Google, or in the pingbacks here.