When animals are created and destroyed solely for a purpose attributed to them by their human overlords, that reduces their utilisable preferences to zero or near zero. Unless a meat producer had reason to believe that inflicting pain on an animal improved the resulting meat product, that pain would almost certainly be a by-product of whatever the farmer chose rather than an exclusive intent. I personally know no farmers that inflict 'pointless' injury on their livestock.

Given any amount of suffering in the animal stock needed to feed, say the US compared to a zero amount of suffering of the in-vitro meat needed to feed the US, if we were basing decisions solely on the ethics of the situation the choice would be clear-cut. As it stands it is simply one amongst many trade-offs, the numbers and data of which I agree would be laborious to define.

The inability to communicate or even experience a preference for the concept of non-existence compared to an experienced or ongoing pain does not invalidate the experience of the pain. In this field of thought I am happy to start from a non-rigorous framework and then become more so if needs be. At a simple level, my model says [for SolvePorkHunger: 'no pig' > 'happy pig + surprise axe' > 'sad pig + surprise axe'].

The practical ways to improve such lives as already exist are, broadly speaking, answered by practitioners of veganism, vegetarianism, cooperative existence with animals (raising chooks, goats for milk, etc etc).