"Replication is hard" can be used to lazily dismiss any and all psychological phenomena; It's a non-argument.Why do you think Dunning-Kruger is so substantial? It's one (and the 1st of that study) psychological study, which is notoriously hard to study because there's a massive replication crisis in that field. Anything in psychology should be taken with massive grains of salt.
If it had never been replicated, and attempts had consistently gone counter to prediction, that would be another thing. But that didn't happen.
As has already been covered, the 'same study' wasn't done with random numbers, because that's very obviously impossible.Just redoing a study with random numbers and getting the same results is not debunking something in your world?