[Python] Optimize collection serialization protocol by homogenization

[chaokunyang]

Feature Request

The new collection serialization protocol in https://fury.apache.org/docs/specification/fury_xlang_serialization_spec/#list can be 2x faster than current one in pyfury. #923 has proved it. we should implement this protocol

Is your feature request related to a problem? Please describe

No response

Describe the solution you'd like

https://fury.apache.org/docs/specification/fury_xlang_serialization_spec/#lis

Describe alternatives you've considered

No response

Additional context

https://github.com/apache/fury/pull/923