[FEATURE] Updated Single Node Results

- The spatial bench docs still reference results from ~11 months ago
- The CI runner has recent changes in expected output but its not specified whether these are supposed to be carried over to official benchmarks as well
- I've been running my own spatial bench harness for cross-engine comparison [harness](https://github.com/pranav-walimbe/PyCanopy/tree/main/bench/spatial_bench) for benchmarking pycanopy but wanted to make sure I'm making the right comparisons since I'm aligning with changes to CI runner + huggingface data copied to S3 as source
- Interested in either seeing new official results (perhaps with PyCanopy) included or just getting confirmation that my harness aligns with spatial bench rules