I'm not denying any of that. What I'm saying is that when you randomly give a person a weak cocktail of random alkaloids and then ask them if they feel better, you're going to get a very unscientific mix of placebo and outright false information.

When you repeat such an experiment on large sample sizes with no control over the other myriad of environmental influences on the subjects, even after attempting to control for confounding factors you're still going to end up with extremely noisy data made effectively useless by just as many contradicting studies which find no effect. You see it all over the place - eggs and cholesterol, coffee harm/benefit, wine harm/benefit. These studies are all intimately highly flawed because they are empirical soft sciences with very little control over the large number of chaotic interactions among and within their subjects.

So when people say things like "drink coffee and eat lettuce to control inflammation during COVID infection" without a disclaimer, they're being [unknowingly] irresponsible, to say the least. Especially considering the dose of active compound in something like lettuce is likely to be totally insignificant.