I'm not sure I agree with that. If you publish your methods and make your data available, then a replicator has all they need. They'd have to reimplement the data pipeline anyway, otherwise it's not a replicated result it's just someone running your notebook with potentially the same bugs in it again.