How is it not falsifiable? They found a correlation between susceptibility to bullshit and the result of a previously established cognitive tests.

> Imagine if I said "People who skip breakfast are worse at their jobs". Its so vague, its always true.

That's a terrible example of your point. As long as you can define a metric for "worse at their jobs" (it'll vary a ton based on which job we're talking about, but it still sounds like something you could assign a metric to) then you have a really clear and testable hypothesis.