Create a runnable benchmark script that reproduces the performance numbers in benchmarks/README.md.
Should cover:
- Scaling from 100 to 10,000 vertices
- Incremental write latency measurement
- Full verification latency measurement
- Output: CSV or JSON with raw timing data
Create a runnable benchmark script that reproduces the performance numbers in benchmarks/README.md.
Should cover: