[FEA] Exploring iceberg statistics improve planning.

[liurenjie1024]

Iceberg has richer statistics:

https://iceberg.apache.org/docs/nightly/spark-procedures/#table-statistics
https://iceberg.apache.org/docs/nightly/spark-procedures/#partition-statistics

We could also consider rewrite data files so that it's better sorted:

https://iceberg.apache.org/docs/nightly/spark-procedures/#rewrite_data_files