Use bulk insert for the first write to an empty Hudi table

For the first batch to an empty table, given no tagging/indexing is needed, we can always use bulk insert (regardless of what's specified as operation type such as INSERT, UPSERT) to improve the write performance with reasonable file sizing.

## JIRA info

- Link: https://issues.apache.org/jira/browse/HUDI-5546
- Type: Improvement

The issue names no files, tests, or entry points. Start by locating the empty-table write path and existing bulk-insert handling; done means the first batch uses bulk insert when the table is empty and no tagging or indexing is needed, regardless of the requested operation type, with the expected file sizing and performance preserved.

Written by the indexing model from the issue text.