[Performance] Compare Collection performance benchmarks across components
We now have performance numbers for common collection operations in the following components:
- API server
- Python SDK
Now that we have these numbers, the next step is to compare them and look for places where there are substantial, achievable performance improvements. We should be able to prioritize improvements based on which ones provide the most clock time savings based on development effort.
The outcome of this story should be implementable stories, with information about the expected time savings for each.