Add system audio as a Python source

[Raphjacksun7]

Summary

• add Source.system_audio() to the Python Session API
• reopen discovered system-mix entries through the same Source declaration
• bind Python 0.1.4 to the published PocketStation Core 1.1.9 package
• document the choice between one application and the complete desktop output mix

Verification

• native tests and strict native Clippy
• Ruff lint and formatting
• strict MyPy
• 275 Python tests
• registry-locked wheel build and isolated consumer
• registry-locked source archive build and isolated consumer

Core 1.1.9 captured controlled audible system output at 10 ms and 20 ms on macOS, Linux, and Windows. The Windows and Linux VM results establish platform correctness, not physical-device or latency performance.