Seems like the perfect recipe for publication bias. Give scientists the freedom to decide when they publish, so if they don't like the result they can just publish "never".

In other words: If you want unreproducible crap science to continue just go ahead.