It seems to me that this logic necessarily makes you choose torture over dust specks, since one can construct a nearly continuous sequence of animals between chickens and humans and patch the gaps, if any, with probabilities. But you are probably OK with that, since you write

While I'm uncertain where along there things start getting up to significant levels, I think it's probably somewhere that includes no or almost no animals but nearly all humans.

and this argument breaks down once you start making comparisons like "0.01% odds of one human dying now vs all animals dying now" or "1 day reduction in the life expectancy of one human vs all animals dying now" etc.