perf: reuse decoded optimized query cache entries

[fresioAS]

Description

Avoids reading and decoding the same optimized-query cache entry twice during model schema propagation.

Previously, schema processing loaded each optimized-query cache entry in a worker process to calculate the model hashes and mapping schema. After the worker completed, the parent process reopened the same gzip/pickle cache file to populate the local model's query renderer.

This change:

• Returns the already-decoded OptimizedQueryCacheEntry from the schema worker.
• Applies that entry directly to the parent process's model.
• Avoids the second filesystem read, gzip decompression, and pickle deserialization.
• Preserves cache-miss and invalid-cache fallback behavior.
• Preserves the existing optimized query, renderer violations, model hashes, and mapping schema.
• Works with both synchronous and multiprocessing schema execution.

This optimization applies independently of project-index loading. Its live performance impact will be validated separately because it trades the duplicate disk read for transfer of the decoded entry across the worker boundary.

Test Plan

Still waiting to test in live env - will update

Checklist

• I have run make style and fixed any issues
• I have added tests for my changes (if applicable)
• All existing tests pass (make fast-test)
• My commits are signed off (git commit -s) per the DCO