spark_fdw or provide native Spark integration

We could consider writing a spark_fdw (foreign data wrapper) to enable querying data in Spark.

Or we could build a tight integration between Spark and PostgreSQL / Citus. In this scenario, Spark manages distributed roll-ups and PostgreSQL acts as the presentation layer.

Start by comparing the two directions described: a spark_fdw for querying Spark data, or tighter Spark integration with PostgreSQL/Citus. Review the existing Spark and PostgreSQL/Citus integration points before choosing an approach. Done means one selected direction is specified and implemented with a clear query or distributed-rollup workflow.

Written by the indexing model from the issue text.