Add benchmark for executing fields with arguments

[patrick91]

This adds a benchmark for executing fields that take arguments.

None of the existing benchmarks resolves a field with arguments, so argument coercion (get_argument_values → coerce_argument → coerce_input_literal) isn't tracked by CodSpeed at the moment. The new benchmark resolves a list of 1,000 objects whose field takes a literal argument:

query ($count: Int!) { items(count: $count) { value(multiplier: 2) } }

For context, we noticed this while upgrading Strawberry to 3.3.0: on an equivalent 2,000-field query, plain graphql-core execution went from about 5.5 ms on 3.2.11 to about 7.4 ms on 3.3.0 locally, and a good part of the extra per-field work is in argument coercion. Having a benchmark for it should make it easier to track and improve. I have a couple of small optimizations for that path that I'm happy to send as a follow-up PR if they're welcome.

The benchmark follows the style of test_execution_sync.py and passes ruff and mypy.

AI: I've done this with Claude 😊

[codspeed]

Merging this PR will not alter performance

✅ 19 untouched benchmarks
🆕 1 new benchmark

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
🆕	test_execute_field_arguments_sync	N/A	126.9 ms	N/A

---

_{Comparing patrick91:benchmark-field-arguments (d49dddb) with main (894141f)}