The study you cite shows an opposite result.

Are there independent replications that actually show the same effect? I recall an attempted replication that produced equal miscalibrations on the upper and lower ends of the distribution, suggesting that the effect was just mean regression.

Your link also says that the Dunning-Kruger effect does not quite work for Asian cultures. I wonder if these persuasion tricks aren't also tied to culture. Do the tricks work in places/cultures were trust is generally low?