Um, is someone maybe a little too overconfident of their knowledge of the Dunning-Kruger effect?

Instead of that cartoon, perhaps you should find and post some real data.

Your text description is okayish, but the graph is not, and they did not find "the opposite" of low people rating themselves low and high people rating themselves high.

There is an important lesson in their results, but you haven't captured it very well.