Author makes a really good point but I think he presents it in a way that doesn't make it obvious and sets it up as ego stroking "science vs engineering" (which

Author makes a really good point but I think he presents it in a way that doesn't make it obvious and sets it up as ego stroking "science vs engineering" (which I can already see in the comments). Science on it's own is not robust enough to build models for real-world decision making - things we could consider in the domain of engineering. If a scientific paper finds a correlation between red wine and heart disease this may be interesting and important clue for scientific study but this is not enough to develop a model of nutrition and certainly not enough to recommend consuming red wine for hearth health. Plenty of people don't realize this and you see articles "red wine good for your heart" - without a robust model of human metabolism/nutrition we just can't make claims like this. The food pyramid is a perfect example of this flaw.