Trying to understand statistical methods through the lens of replication

Editorial note: Erik has stated the problem extremely clearly here and on another thread, so we won’t be debating the replication probability formula again here.