[Refactor] Migrate to DataUpdateCoordinator for centralized polling

[SecKatie]

Summary

Migrate the integration from individual entity polling (SCAN_INTERVAL + async_update) to Home Assistant's recommended DataUpdateCoordinator pattern for centralized, efficient data fetching.

Current Implementation

The integration currently uses the legacy polling pattern with:

• SCAN_INTERVAL defined per platform (e.g., 30s for lights/switches, 10s for locks)
• Individual async_update() methods on each entity
• Callback functions for push updates (async_update_callback)

This results in:

• Multiple API calls when many entities of the same type exist
• No automatic deduplication of requests
• Manual error handling per entity

Proposed Changes

Adopt the DataUpdateCoordinator pattern as documented in Home Assistant Developer Docs.

Benefits

• Single API call fetches data for all entities of a type
• Automatic optimization - only polls when entities are listening
• Built-in error handling with retry logic and rate-limit backoff
• Reduced state writes when data hasn't changed (always_update=False)
• Cleaner code - entities extend CoordinatorEntity for automatic updates

Implementation Steps

1. Create coordinator classes (e.g., WyzeLightCoordinator, WyzeSwitchCoordinator)
2. Override _async_update_data() to fetch device data from wyzeapy
3. Use async_config_entry_first_refresh() during setup
4. Migrate entity classes to extend CoordinatorEntity
5. Remove SCAN_INTERVAL and individual async_update() methods
6. Keep existing callback mechanism for push-based updates where applicable

Platforms to Migrate

• light.py
• switch.py
• lock.py
• sensor.py
• climate.py
• camera.py
• binary_sensor.py
• alarm_control_panel.py

References

• Home Assistant: Fetching Data
• CoordinatorEntity Documentation