How interesting would it be if every PhD thesis had to have a "replication" section, where they tried to replicate some famous paper's results.