test(lint): replace the last flaky wall-clock ratio gate

[takeokunn]

lint-introspection's the_cost_of_each_rule_grows_linearly_with_the_input asserted a wall-clock doubling ratio below 3.0. It failed three CI runs on untouched code — at 4.30× and 9.24× — and passed 3/3 on re-run at load average 69. It was noise on PRs #90, #91 and #92.

Its own docstring claimed "neither is a wall-clock threshold — those are what make a test flaky on a loaded machine", which was wrong about itself. A ratio of two wall-clock measurements normalizes for machine speed but not for load changing between the two measurements, which is exactly what a shared box does.

The split

Following the precedent PR #85 set for lint-documentation, lint-contract-annotation and lint-performance:

• each_rule_is_dispatched_once_per_head_match keeps the deterministic half — the engine's own invocation counter — and gains an explicit assertion that doubling the definitions doubles the dispatch count. That is the shape the ratio existed to catch (a rule re-walking the file per match), stated as a property of dispatch rather than of nanoseconds.
• ignored_bench_doubling_ratio keeps the timing table behind #[ignore], with the invocation to run it by hand.

Mutation-checked: perturbing one expected invocation count by 1 fails the new test, so it still discriminates rather than passing vacuously.

Audit of the rest

No others need fixing:

site	state	why
lint-form-shape	already #[ignore]d	PR #85
lint-performance	already #[ignore]d	PR #85
lint-package-hygiene	left alone deliberately	normalizes against a control rule measured in the same pass, so a loaded machine moves both ratios together; its comment says exactly this, its bounds are 9.0 absolute and 2.2× control, and it has never flaked

That last one is the technique worth keeping: a bare constant can't distinguish "the rule got slower" from "the machine got busier", but a control measured in the same pass can.

Verification

cargo test -p paredit-feature-lint-introspection → 0 (114 passed, 1 ignored), cargo fmt --all --check → 0, cargo clippy -p … --all-targets -- -D warnings → 0.