Not even that is enough because you might end up with the filling drawer problem. For example you might end up with 50 people trying to replicate it, 49 failing and 1 "succeeding" by chance. The 49 won't publish because they probably got the "wrong" answer by chance, and they don't want to publish negative results, so they will leave it in the filing drawers.
The one that did "replicate" it will publish and then in round 2 there is even more pressure not to publish negative results because "hey, this was replicated before already"
The one that did "replicate" it will publish and then in round 2 there is even more pressure not to publish negative results because "hey, this was replicated before already"