feat: enable batch for instrumentation

[jeanscherf]

Description

This PR changes the default to disable_batch=False (async BatchSpanProcessor) and exposes the parameter so users can override it when needed.

auto_instrument was hardcoding disable_batch=True, forcing SimpleSpanProcessor which exports spans synchronously on the request thread. With concurrent workloads, the network round-trip to the OTEL collector adds up directly on the hot path. A user reported response times dropping from ~40s to ~5s under 9 concurrent requests after patching this locally.

Related Issue

Closes #40

Type of Change

• Bug fix (non-breaking change that fixes an issue)
• New feature (non-breaking change that adds functionality)
• Documentation update

How to Test

1. Call auto_instrument() and send concurrent requests to an instrumented agent exporting to an OTLP endpoint
2. Verify response times are not impacted by span export latency
3. Verify auto_instrument(disable_batch=True) still works for cases where synchronous export is needed (e.g. short-lived scripts)

Checklist

• I have read the Contributing Guidelines
• I have verified that my changes solve the issue
• I have added/updated automated tests to cover my changes
• All tests pass locally
• I have verified that my code follows the Code Guidelines
• I have updated documentation (if applicable)
• I have added type hints for all public APIs
• My code does not contain sensitive information (credentials, tokens, etc.)
• I have followed Conventional Commits for commit messages

Breaking Changes

None. The new default (disable_batch=False) is a behavior improvement. Users who relied on the previous synchronous behavior can restore it with auto_instrument(disable_batch=True).

Additional Notes

Trade-offs of BatchSpanProcessor (the new default):

• Spans in the queue are lost if the process is killed abruptly (SIGKILL)
• In short-lived scripts or serverless runtimes, call tracer_provider.force_flush() on shutdown to avoid losing the last batch
• Export errors surface in a background thread (logged, not propagated)

For these cases, disable_batch=True remains available and is documented in the user guide.

[LucasAlvesSoares]