I'd say that I have more confidence in the Redshift query planner and execution engine than that of presto. And having the option to join that with data stored

I'd say that I have more confidence in the Redshift query planner and execution engine than that of presto. And having the option to join that with data stored in-memory on redshift nodes is very attractive.