In my former career as a broadband network performance specialist it turned out that on close examination there was no science to trust. It was a pile of numerate craft with basic failures to understand cause and effect. The most elementary aspects of engineering (like the ability to quantify demand and supply and predict failure modes) were missing. So given that experience — a whole industry with insane beliefs are irrational economics — I have no reason to imagine it’s a better situation in other comparable human endeavours like medicine or energy.