Batch modularize benchmark and chunk search tests

[yvette-carlisle]

Summary

• Split chunk search acceptance tests into scenario modules and relation-context fixtures.
• Split real-world benchmark core assertions into smoke, capture/source, blocker, and adversarial scoreboard modules.
• Split root aggregate summary assertions into count, scoreboard, and suite summary modules.

Validation

• RUSTC=/run/current-system/sw/bin/rustc cargo make fmt-rust && RUSTC=/run/current-system/sw/bin/rustc cargo make check (after chunk search split; 403 passed, 92 skipped)
• RUSTC=/run/current-system/sw/bin/rustc cargo make fmt-rust && RUSTC=/run/current-system/sw/bin/rustc cargo make check (after benchmark/root aggregate split; 403 passed, 92 skipped)