This isn't politics, it's safety. I accept the argument that the CDC has become overly innumerate in how to live a healthy life, but it's not a liberal or conservative idea to be cautious.
Many people seem to unrealistically demand scientific recommendations to be "guarantees." However, scientific conclusions are very precisely made best guesses, built on humanity's knowledge.
It's good that the supply chain of science is traceable, but it doesn't guarantee truth by any means. Their are many non-scientific human endeavors that have traceability, take software development for instance.
I think the distrust comes from many fields having fairly obvious political trends, in particular social sciences. Another source of distrust is overextrapolation of science by authorities. For instance, science might say that certain drugs are harmful if abused, and politicians may use that to claim the war on drugs is based on science, which obviously is not true. Lastly, some fields struggle producing consistent and falsifiable results, such as economics and psychology.
58% (quite high, higher than a lot of biology related field) are positive replications.
But! some of those use the same datasets, with different time bases, and most are narrow replication of high impact studies.
Also, economics now is not a singular field, and I'd bet gp was talking about macroeconomics. You could be more charitable.
You can get a pretty good estimate of how many lives would be saved if we smoked less, or drank less, or did any number of other things, but ultimately that's an input into a public policy decision making system that doesn't have a provably right answer unless everyone involved agrees on what the desired outcome is.