Swap JDBC for Arrow ADBC

[jaidisido]

Leveraging the Arrow ADBC Driver has a number of advantages:

• A single, universal driver would replace all existing, proprietary drivers (mssql, postgres...)
• Better performance due to not processing each row separately and tapping into pyarrow to avoid serialization/deserialization

Additional context
https://arrow.apache.org/docs/dev/format/ADBC.html

P.S. Please do not attach files as it's considered a security risk. Add code snippets directly in the message body as much as possible.

[LeonLuttenberger]

The implementation for this is in progress for Postgresql (see development branch). However, we are currently blocked by this bug in Pandas: pandas-dev/pandas#57539