What would happen when running mempool web app on top a catched up pruned node and track back to old transactions no longer in persistent storage?
Would it behave buggy? If so, would be in scope not to be buggy? In other words, does this software require to be on a full node?
Don't have any pruned node at hand to check.