Optimization: Async cleanup of connections before pooling

[jduo]

Rework the connection state clean-up to happen asynchronously instead of blocking the application thread.
This provides similar performance benefits to skipping the UNWATCH entirely, but is safer due to not
changing the semantics and also more applicable to scenarios where there actually are WATCHed keys.

• You have read the Spring Data contribution guidelines.
• You use the code formatters provided here and have them applied to your changes. Don’t submit any formatting related changes.
• You submit test cases (unit or integration tests) that back your changes.
• You added yourself as author in the headers of the classes you touched. Amend the date range in the Apache license header if needed. For new types, add the license header (copy from another file and set the current year only).

[jduo]

This yielded about a 1.75x performance improvement compared to the un-optimized version when testing with the benchmark and server running locally.

It would yield a greater benefit if real network is involved since this lets us do work while a network roundtrip is happening rather than just waiting.