Hey @Mike
The concept of replicated tables is it is distributed across all distributions.
So there is no specific need to do anything once you pause and resume ideally.
But you can create statistics if needed but ideally for replicated tables I would suggest those are not needed.
let synapse create them automatically if needed