Create an Official Beam I/O Connector for Cloud SQL

### Feature Description

The current Cloud SQL Python connector is great. But in it's current form, when used via Apache Beam/Google Cloud Dataflow, it is not recognized as an official Apache Beam framework construct. This can lead to unbalanced workers in Dataflow when dealing with any kind of volume in a pipeline.

I suggest creating an official [Beam I/O Connector](https://beam.apache.org/documentation/io/connectors/) so that Cloud SQL can have optimized usage with Beam/Dataflow.

### Sample code

_No response_

### Alternatives Considered

There is current [non-native Cloud SQL](https://github.com/jccatrinck/dataflow-cloud-sql-python) support, but it requires a custom container. This can cause slowdowns and longer approval processes at regulated customers. An official native I/O connector would be ideal.

### Additional Details

_No response_