Implement Databricks-to-HAPI writeback path

[slusset]

Summary

• change the Databricks writeback flow so DLT prepares rows and a notebook task performs real HAPI FHIR writes through the orchestrator
• add an execution notebook for batch writeback and upgrade dead-letter reprocessing from placeholder logic to actual retries
• update the Databricks bundle and docs for explicit FHIR endpoint configuration and real outcome logging

Verification

• uv run ruff check src/ tests/
• uv run ruff format --check src/ tests/
• uv run pytest tests/unit -v --tb=short
• uv run --group docs mkdocs build --strict
• databricks bundle validate -t dev --profile dbc-155b9d6f-de4d --var="catalog=ehr_writeback_dev"

Runtime follow-up

• rerun the deployed Databricks jobs in the workspace and verify writeback_log status values, HAPI resource creation, and rerun idempotency evidence