That is the problem: Why wouldn't try to do a replication?

Because the incentives are to provide "new research", not to confirm or deny a known result.

It seem to me that most probably no-one has gotten tenure by reproducing, disproving results.