Also, Presto and Databricks(Spark) is just a layer on top of other storagea, it cannot optimize storage for querying, as you do storage yourself.