A huge benefit of single-database operations at scale is point-in-time recovery for the entire system thereby not having to coordinate recovery points between data stores. Alternatively, you can treat your queue as volatile depending on the purpose.