Alloy MVC and Titanium.HTTPClient: Concurrent Sync Request Latency

Integrating the Alloy MVC data layer with Titanium.HTTPClient involves delegating REST operations to the native networking stack. When executing multiple collection.fetch() or model.save() calls simultaneously, the SDK manages multiple concurrent HTTPClient instances.

There is uncertainty regarding how the Titanium SDK handles the queueing of these parallel requests at the bridge level. It is unclear if the framework implements a built-in concurrency ceiling or a specific backoff policy when the underlying native OS socket limits are reached during heavy synchronization loads.

Assuming a recent Titanium SDK version, what is the documented behavior for request prioritization during concurrent Alloy sync operations? Does the SDK provide a mechanism to limit the number of simultaneous active HTTPClient instances to prevent latency spikes?