I don't think that D–K really has much to do with it: as originally presented, it predicts that the worst performers tend to think of themselves as close-to-average performers, not as the best performers. (In particular, there's no "U curve" of self-assessment.) Imitation of expertise is more of a kind of personal bravado.