Apply buffering changes to Databricks Destination

[ChristopheDuong]

Tell us about the problem you're trying to solve

As part of #10260, refactor destination-databricks to use serialized buffering strategy.

[tuliren]

Databricks is going to roll out their internal staging solution within the month. Ideally, we should wait until their solution is ready, and do the two changes together:

• Migrate our S3 staging loading method to use the serialized buffer.
• Add a new loading method to use Databricks' internal staging.

If we are going to do them together, please do not assign this to GL. We should work on this, because there will be some exploratory work with adopting Databricks' solution.

[martindlarsson]

Any news on the Databricks solution? Is there a ticket for this I can follow?