Well, of course, because the terms 'handful,' and 'overly large' are undefined. In the article's case, it refers to 1% causing 9%. Disproportionate, but far from a majority.

Contrast with threads about the '80/20' rule. In those, people start saying something not much different, like:

> I suspect the sentence "80% of <effects> come from 20% of <causes>," is true for all groups.

We'll always have a distribution of a range of behaviours, and we'll always have cognitive bias; these in combination will lead to any number of reasonable 'x of y' rules about populations.