Fun Fact: Rat Park has never replicated, and is now considered bunk: https://en.wikipedia.org/wiki/Rat_Park#Replication

One graduate student failing to replicate does not support the assertion that it 'is now considered bunk'

You’re being disingenuous. ”There has been little subsequent interest in replicating the studies due to several methodological issues present in the originals”

A paper with many egregious methodological issues is, at worst, a paper that should be ignored, as though it never happened. We cannot say that it is "wrong" or "incorrect" or "bunk" because it's a nothing, a non-argument. Until the experiments are repeated, sans egregious methodological issues, we can't logically make that determination.

So if it’s generally accepted that the world is flat, we shouldn’t listen to anyone claiming otherwise?

A better analogy would be "If it’s generally accepted that the world is flat, we shouldn't assume its round because one student was unable to measure the curvature."

The machine refuses to work against its self-preservation.

Plato wrote in his famous work on democracy that the body of citizens in a polis should not exceed 10000, else the direct democracy disintegrates, and both electing officials consciously, and keeping them accountable becomes hard.

The representative democracy in the US works with 350M people, 240M eligible to vote, is hopelessly part these confines. It's not a wonder that it works not really ideally; it's a wonder that it works at all for nearly 250 years.

The US needs to upgrade the mechanisms of its democracy to better suit its scale. Switching to Condorcet voting or other preferential voting schemes on municipal and state levels in a few states is one such upgrade. Maybe people will start electing more officials who actually care, and who might shake the inadequate equilibria, in drug-related policies, and elsewhere.