It's mostly not anything specific to going vegan. Just the general truism that effort used for one purpose could be used for something else instead. (Plus I sometimes donate extra precisely for the purpose of "offsetting", which I wouldn't otherwise be motivated to do.)

Mostly just changing old habits, plus some anticipated missing of distinctive desired tastes. It's not an unreasonable ask or anything, but I'd much rather just donate more. (In general, I suspect there's insufficient social pressure on people to increase our donations to good causes, which also shouldn't be "so effortful", and we likely overestimate the personal value we get from marginal spending on ourselves.)

I don't understand the relevance of the correlation claim. People who care nothing for animals won't do either. But that doesn't show that there aren't tradeoffs in how to use one's moral efforts on the margins. (Perhaps you're thinking of each choice as a binary: "donate some" Y/N + "go vegan" Y/N? But donating isn't binary. What matters is how much you donate, and my suggestion is that any significant effort spent towards adopting a vegan diet might be better spent on further increasing one's donations. It depends on the details, of course. If you find adopting veganism super easy, like near-zero effort required, then great! Not much opportunity cost, then. But others may find that it requires more effort, which could be better used elsewhere.)

My main confusion with your argument is that I don't understand why donations don't also count as "personal ethics" or as "visible ethical action" that could likewise "ripple outward" and be replicated by others to good effect. (I also think the section on "equity" fundamentally confuses what ethics should be about. I care about helping beneficiaries, not setting up an "equitable moral landscape" among agents, if the latter involves preventing the rich from pursuing easy moral wins because this would be "unfair" to those who can't afford to donate.)

One more specific point I want to highlight:

...where harm is permissible as long as it’s “offset” by a greater good

fwiw, my argument does not have this feature. I instead argue that:

(1) Purchasing meat isn’t justified: the moral interests of farmed animals straightforwardly outweigh our interest in eating them. So buying a cheeseburger constitutes a moral and practical mistake. And yet:

(2) It would be an even greater moral and practical mistake to invest your efforts into correcting this minor mistake if you could instead get far greater moral payoffs by directing your efforts elsewhere (e.g. donations).

Just to clarify: Spears & Geruso's argument is that average (and not just total) quality of life will be significantly worse under depopulation relative to stabilization. (See especially the "progress comes from people" section of my review.)

The authors discuss this a bit. They note that even "higher fertility" subcultures are trending down over time, so it's not sufficiently clear that anyone is going to remain "above replacement" in the long run. That said, this does seem the weakest point for thinking it an outright extinction risk. (Though especially if the only sufficiently high-fertility subcultures are relatively illiberal and anti-scientific ones - Amish, etc. - the loss of all other cultures could still count as a significant loss of humanity's long-term potential! I hope it's OK to note this; I know the mods are wary that discussion in this vicinity can often get messy.)

I wrote "perhaps the simplest and most probable extinction risk". There's room for others to judge another more probable. But it's perfectly reasonable to take as most probable the only one that is currently on track to cause extinction. (It's hard to make confident predictions about any extinction risks.) I think it would be silly to dismiss this simply due to uncertainty about future trends.

What reason is there to think that demographic trends will suddenly reverse? If it isn't guaranteed to reverse, then it is an extinction risk.

I'd guess that (for many readers of the book) less air travel outweighs "buying more" furniture and kids toys, at least. But the larger point isn't that the change is literally zero, but that it doesn't make a sufficiently noticeable change to near-term emissions to be an effective strategy. It would be crazy to recommend a DINK lifestyle specifically in order to reduce emissions in the next 25 years. Like boycotting plastic straws or chatgpt.

Updated to add the figure from this paper, which shows no noticeable difference by 2050 (and little difference even after that):

As a general rule, it isn't necessary to agree on the ideal target in order to agree directionally about what to do on present margins. For example, we can agree that it would be good to encourage more effective giving in the population, without committing to the view (that many people would "personally disagree" with) that everyone ought to give to the point of marginal utility, where they are just as desperate for their marginal dollar as their potential beneficiaries are.

The key claim of After the Spike is that we should want to avoid massive depopulation. Whether you'd ideally prefer stabilization, gradual population growth, or growth as fast as we can sustainably maintain without creating worse problems, isn't something that needs to be adjudicated -- and in fact seems a distraction from the more universally agreeable verdict that massive depopulation is bad and worth avoiding.