Add basic JSONBench-based benchmark

[AdamGS]

Summary

This PR adds a very naive "compress jsonbench as strings" benchmark. The use of vortex_json required some API changes to the Dataset API, but I think this is relatively straight forward.

This is a step towards more variant-focused benchmarks and functionality, trying to get the boilerplate out of the way before I try and land bigger changes.

This PR builds on top of #8339, which includes some missing functionality for vortes-json.

[codspeed-hq]

Merging this PR will degrade performance by 10.62%

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

⚡ 2 improved benchmarks
❌ 6 regressed benchmarks
✅ 1519 untouched benchmarks

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	chunked_bool_canonical_into[(1000, 10)]	31.9 µs	46.8 µs	-31.82%
❌	Simulation	decompress_rd[f64, (100000, 0.0)]	845.8 µs	1,024.6 µs	-17.45%
❌	Simulation	chunked_varbinview_into_canonical[(1000, 10)]	177.2 µs	213.3 µs	-16.95%
❌	Simulation	decompress_rd[f32, (100000, 0.0)]	499.5 µs	587 µs	-14.91%
❌	WallTime	cuda/bitpacked_u8/unpack/3bw[100M]	300 µs	349.9 µs	-14.25%
❌	Simulation	chunked_varbinview_canonical_into[(100, 100)]	274.7 µs	309.5 µs	-11.26%
⚡	Simulation	decompress_rd[f64, (100000, 0.01)]	981.6 µs	846 µs	+16.03%
⚡	Simulation	decompress_rd[f64, (100000, 0.1)]	981.5 µs	846 µs	+16.02%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing adamg/compress-json-bench (1e1a170) with adamg/json-arrow-export (5430d3a)}