Add Criterion benchmark for error conversion paths

[RAprogramm]

Summary

• add a Criterion benchmark that exercises Context::into_error metadata paths and ProblemJson::from_app_error
• document the benchmark workflow in the README and template while bumping the crate to version 0.24.2
• record the change in the changelog and register the bench in Cargo.toml

Testing

• cargo +nightly fmt --
• cargo +1.90.0 build --all-targets
• cargo +1.90.0 test --all
• cargo +1.90.0 clippy -- -D warnings
• cargo +1.90.0 doc --no-deps
• cargo +1.90.0 bench --bench error_paths
• cargo audit
• cargo deny check

---

https://chatgpt.com/codex/tasks/task_e_68d72fd6ac20832b96e28ae91b42335a