refactor: unify model_type semantics by introducing sub_type field

This commit resolves the semantic confusion around the model_type field by
clearly distinguishing between:
- scanner_type: architecture-level (lora/checkpoint/embedding)
- sub_type: business-level subtype (lora/locon/dora/checkpoint/diffusion_model/embedding)

Backend Changes:
- Rename model_type to sub_type in CheckpointMetadata and EmbeddingMetadata
- Add resolve_sub_type() and normalize_sub_type() in model_query.py
- Update checkpoint_scanner to use _resolve_sub_type()
- Update service format_response to include both sub_type and model_type
- Add VALID_*_SUB_TYPES constants with backward compatible aliases

Frontend Changes:
- Add MODEL_SUBTYPE_DISPLAY_NAMES constants
- Keep MODEL_TYPE_DISPLAY_NAMES as backward compatible alias

Testing:
- Add 43 new tests covering sub_type resolution and API response

Documentation:
- Add refactoring todo document to docs/technical/

BREAKING CHANGE: None - full backward compatibility maintained

py/services/model_scanner.py

@@ -275,9 +275,16 @@ class ModelScanner:
         _, license_flags = resolve_license_info(license_source or {})
         entry['license_flags'] = license_flags
 
+        # Handle sub_type (new canonical field) and model_type (backward compatibility)
+        sub_type = get_value('sub_type', None)
         model_type = get_value('model_type', None)
-        if model_type:
-            entry['model_type'] = model_type
+        
+        # Prefer sub_type, fallback to model_type for backward compatibility
+        effective_sub_type = sub_type or model_type
+        if effective_sub_type:
+            entry['sub_type'] = effective_sub_type
+            # Also keep model_type for backward compatibility during transition
+            entry['model_type'] = effective_sub_type
 
         return entry