Address Flake in test_workflow_history_info

[VegetarianOrc]

Add polling for expected history before sending final signal in test_workflow_history_info

What was changed

Wait for the first signal's timers to be committed so the next signal creates a post-timer workflow task with updated workflow.info().

Why?

This avoids a race where both signals are accepted before the worker processes the first one and both make it into the same activation. If that occurs, the query will have the stale history that the final signal is intended to avoid.

Checklist

1. How was this tested?

I created a repro locally that:

1. started the workflow
2. shut down the worker so nothing was polling
3. Sent both signals ensuring they'd be batched together
4. started a new worker and performed the query

After reproducing that way, I was able to confirm the fix that waiting for the signals to appear in history before sending the second signal addressed the issue.