I think you are missunderstanding this a bit, hack has async io. If you have only 1 external system that both indexes the data and stores it, this won't help y

I think you are missunderstanding this a bit, hack has async io. If you have only 1 external system that both indexes the data and stores it, this won't help you that much. But if you have an indexing system and another, with many read slaves, that stores the data, then async will help you. You will do 1 external io request to retrieve a list of ids/identifiers then as many replicas paralel requests to retrieve and format the data.