Add `--disable-echo-cancellation` CLI option and update audio pipeline to support toggling echo cancellation and VAD

Enable debug mode with global configuration and detailed task logging when active

Refactor `main.py`: remove redundant code, modularize functionality, and set multiprocessing to `spawn` for PyInstaller compatibility.

Update `main.py`: modify `uvicorn.run` to use updated `voice_dialogue.api.app` module path.

Enhance `main.py`: add `multiprocessing.freeze_support()` for compatibility and improve `lib_path` existence check before modifying sys.path.

Rename 'src/VoiceDialogue' to 'src/voice_dialogue'.