I concur, but where I disagree is the sentiment "stop relying on others to tell you what to think!"

In order to fix errors and optimize, we need information. Trying to reason about a problem without adequate information is like trying to fix a bug when the only information you have is your QA folks coming back and saying "There's a bug."

but why should they 'eat a tiny bit of everything'?

Suppose you don't know whether something is good or bad for your health. Other people are eating it, so it's probably not insta-deadly, but you suspect that eating too much of it might be bad in the long term. Meanwhile, some people are telling you that it's actually essential for good health (e.g. cholesterol), and some are even worshiping it as some kind of superfood (e.g. wine). Food science is a mess.

So you eat a little bit of it just in case it's essential for good health, but not too much of it in case it turns out to be bad in large quantities. Basically, you hedge your bets.

Wash and repeat for nearly every type of food that exists.