Improve performance by processing style layers in parallel

[pramilk]

Change

Style processing (createStyleLayer method) is one of the most expensive operations performed during initial load and also during style change. This takes anywhere from 100-400ms, depending on machine speed, number of layers, expressions used in style etc.

• Currently: Styles are validated (If validation is enabled) --> Styles are processed on main thread (via createStyleLayer function) --> broadcasted to all worker threads --> Styles are processed on each worker thread (via createStyleLayer function).

• New change: Styles are validated (If validation is enabled) --> broadcasted to all worker threads --> Styles are processed in parallel (via createStyleLayer function) on both main and worker thread.

Note: In our observation doing a prewarm or an external worker as per this PR (#2130), gives much better results as worker threads are ready by the time it gets the broadcast message to truly allow parallel processing.

Launch Checklist

• Confirm your changes do not include backports from Mapbox projects (unless with compliant license) - if you are not sure about this, please ask!
• Briefly describe the changes in this PR.
• Link to related issues.
• Include before/after visuals or gifs if this PR includes visual changes.
• Write tests for all new functionality.
• Document any changes to public APIs.
• Post benchmark scores.
• Add an entry to CHANGELOG.md under the ## main section.

[HarelM]

Can you please add a test to test that somehow?

[pramilk]

Can you please add a test to test that somehow?

Added some tests. Also while testing realized that "custom" filter is not required as a style Json does not support it. Added some more tests for it.

[HarelM]

THANKS!