I think there is something of a threshold for "liar", i.e. the implication is that it's habitual, or that the consequences are serious (sometimes, but not always both of those). And there is a huge variation in the trustworthiness or lie-frequency of individuals, that I think would be fairly consistent if you watch the same person over time, then compare them to others. A reasonable person has probably spotted high frequency liars at various points of daily life and made mental notes not to trust them, or the opposite for the opposite traits, and made these determinations correctly.

So I think the point is a little facetious.