> At some point with this line of reasoning you fall into the precautionary principle
Is there something wrong with the precautionary principle? It would be one thing if the supposed benefit was something really incredible, like extending lifespan or curing cancer. Then maybe we should be less cautious. But we're talking about adding something to your ice cream to make it look nice for longer.
> it's about people eating hyperpalatable low-satiety packaged food, which are obesogenic. Getting rid of carrageenan does precisely nothing to address that problem; getting rid of cane sugar, which is not a UPF ingredient, absolutely does.
I'm not advocating anyone eat tons of sugar, but sugar consumption and the obesity epidemic are not very well correlated. I would agree that we should eliminate "hyperpalatable low-satiety packaged food", but if you were to hypothetically ban basically all non-salt/sugar preservatives, stability agents, flavor enhancers, colors etc. then you almost eliminate this entire product category, because it's no longer practical to produce and sell, easy to consume, or as marketable. Even banning corn syrup in packaged food (as a UPF ingredient) would be a positive move because forcing its replacement by cane sugar (regardless of whatever alleged health problem HFCS s may or may not have) would mean that such products become less economically viable, because its more expensive and less stable.
> the naturalist fallacy.
The naturalistic "fallacy" is approximately true for diets. We are animals that evolved in a way that optimized for the consumption of various foods in our environments. We're some of the most wildly complex chemical systems in the world, and we have remarkably broadly adapted digestive systems with a pretty good tolerance, so you can get away with throwing a lot of stuff down the pipe without anything bad happening. But that's exactly why "eat the same foods people always have and not bizarre lab concoctions" is a useful heuristic for health. It's entirely possible that various additives are perfectly fine or even pro-health, but it's not the way to bet as a general principle, and it's impractical to conduct meaningful long-term nutritional studies to find out with any real assurance.