This is a very basic math error. If you multiply two unknown quantities, you can not assume anything about the result, even if one of them is very big. So basically what you're saying is "we have no idea if it's dangerous or not, and there's absolutely no evidence on any danger present, but if it were, you could DIE!!! so PANIC!!!" This would be an extremely irrational and stupid behavior.
>>> But the benefits are small to nonexistent, especially for your typical first-worlder who as I mentioned is drowning in food.
The problem is that it is very fragile. If the anti-science idiots ever prevail, the whole food abundance we have now could very well collapse pretty quickly - without science, malthusian scenarios are very real, and the people who predicted massive hunger by 2000s could very well be right if science did not intervene. If the idiots ever succeed in their war on science, first world would be in no way insulated from the disastrous effects of their victory. The whole abundance is the direct effect of what you call "food liberalism" and it is the only thing maintaining it.
>>> The same calculus underlies decisions to take out insurance policies
Not at all. Insurance companies have very accurate actuarial and statistical data, which they use in very accurate and specific ways to issue insurance. They never base their decisions on "multiply two unknown numbers, make assumptions on the result, panic". Your decision on buying the insurance may very well be irrational, but the insurance market itself is very rational.
>>> Statistically you'll probably be better off not buying insurance
It is not true, since you can not be better "statistically" - the fact that it didn't happen to 100000 people would not improve your situation if it happens to you. The whole function of insurance is to make statistics work for you, without it statistics would be just a useless trivia - just as the fact that if Bill Gates walks into a bar, the average income of the bar visitors jumps - is useless for anybody in the bar, because they get no money from it. Insurance is the way to get you some money from statistics.
The subject is conservatism. A conservative is not someone who rejects all of a thing, but someone who is skeptical and suspicious of new things.
As a result of many past issues with novel food products, people have become increasingly conservative about food. That doesn't mean that they reject all industrial or scientific agriculture, but they're likely to be suspicious of things that haven't been around for more than one or two generations. They're quick to reject novel changes in the food supply and very slow to accept them.
Along with having been burned in the past, I can think of another reason why this might be rational. The cost function is weighted heavily on the conservative side. It feels like something that could be modeled well with game theory.
The cost of food conservatism is likely a small increase in the amount of money paid for food. This is a very small cost for most first-worlders, where food is so available that obesity is a health crisis. The benefit is that the risks, if any, are well-known and are known to be tiny.
The cost of food liberalism, by contrast, is an unknown chance of getting cancer, heart disease, diabetes, allergic reactions and inflammatory disorders, etc. The unknown chance is probably very small, but many of those health effects have astronomically high costs that possibly include death or permanent disability. But the benefits are small to nonexistent, especially for your typical first-worlder who as I mentioned is drowning in food.
The same calculus underlies decisions to take out insurance policies on unlikely events such as floods or tornadoes. Statistically you'll probably be better off not buying insurance -- otherwise how would it be that the insurance company could make money? But you'd rather trade the risk of losing everything you own and having no recourse for a small amount of money that you're not going to miss.