Right, and that choice of axiom is how two reasonable people can come to wildly different conclusions about a subject even after spending time resolving reconcilable differences.

But empirically speaking, almost all people would agree at a high-level that minimizing suffering is an agreeable axiom.

>almost all people would agree at a high-level that minimizing suffering is an agreeable axiom

Yes, they would, but there is a weighting based on the perceived status of the animal. Otherwise we would feel compelled to destroy all carnivores, since a single carnivore can kill many, many prey throughout its lifetime. We agree that the utility a lion gets from eating a dozen zebras is greater than the suffering of those zebras. Also, note that predators, do not just kill strictly what they need for food(biggest example is housecats, though other examples include wolves killing elk https://www.nationalgeographic.com/news/2016/03/160325-wolf-...).

Similarly, humans have a weighting function. Many people think that the utility/pleasure that a human gets from eating octopuses is greater than the suffering of those octopuses.

So if one morality position is animal suffering = bad then rather than arguing using animal well being/suffering, they should instead for example promote artificial meat that look and taste exactly like octopus. That I believe would be more effective approach.

yes so if you can make almost all people would agree that eating octopus is bad or eating octopus is good then that will be the outcome.