[gfodor]

No, I reject the over-simplified model in general, but my argument is that whatever terms you are, in your own model, using to mean "good" and "bad" and "too much" and "too little" are irrelevant to the question of, even if you could define them clearly, if you can also place and measure a valid threshold. Given the problem here specifically is the epistemological value of consuming views you currently disagree with, by setting yourself up as judge and jury of who is considered worth listening to you're just begging the question, regardless of your framework to act as said judge.

At the end of the day, you have scarce attention, so there are necessarily heuristics you need to use to navigate information and consuming it due to opportunity costs. But my argument is that pegging some people and not others as "too skeptical" is low on the list of good heuristics.