It's from a discussion on p-f where someone was inviting unqualified speculation as to what rankings, standards, skill levels, etc. were equivalent to each other, like the ever-popular USPSA B class = IDPA M class comparison.
He then wanted to further relate those demonstrated levels of skill to the concept of 'overlearning' or 'automaticity.' The significance to him was that there is apparently a lot of scientifically sound evidence that a motor skill reaching that level of automaticity, unconscious competence, etc., is subject to considerably less degradation under stress.
Eventually, a near-consensus was reached that whether a person's skill reached the point of automaticity was separate and distinct from any particular skill level, and that while a low level of skill may or may not have reached automaticity, a certain high level of skill probably did indicate automaticity.
Charts were modified and thrown up in total and complete speculation as to what that level of skill might be.