>“We could imagine a significant shift from beef to chicken, and that by itself goes a long way.” (Poultry production has about one-eighth the climate impact of beef production.)

I think the possible benefits of this approach might be exaggerated. The uniquely "large" environmental impact of beef cattle results from methane emissions, but methane is not the big challenge; land use is. While methane is a potent greenhouse gas, it also has a shorter atmospheric lifetime than CO2/N2O, which, to me, suggests that it should be addressed after the other gases; the impact of a reduction in methane emissions will be felt quickly, and the risk of methane accumulation is lower.

Land use -- with 40% of the world's land devoted to agriculture -- is much more of a problem and I see less indication that it will be addressed by changing the relevant animal. A little, sure, but a lot of the land currently used for pasturing cattle isn't appropriate for chickens anyway, and I'm not convinced that it takes 8x as much land to raise a pound of beef vs. chicken.

Lab-grown meat is/will be a huge net win for the environment and hopefully our wallets, and I think it will be the final compromise on this front, at least for the next few hundred years. A few decades of lead time doesn't bother me for reasons discussed above.

On another front, the densest sources of plant protein after soy are hempseed, lentils and pumpkin seeds. These rarely make it into our diets, though. You need more than mere tax incentives -- the strong flavors of those seeds need to be managed somehow. I made a "curry" with a hemp-milk base once. The only thing that corrected the smell was a lot of Worcestershire sauce. Ironically, it's vegetarians who won't eat fermented anchovies.