Reduce repeated work for fields without arguments #298

Executor.execute_field calls argument coercion for every field, including fields with no arguments.
Its async error handler also captures local values during synchronous execution.
This adds repeated work when executing introspection queries on large schemas.

A local change uses an empty dictionary when the schema field and query node both have no arguments.
It passes async error-handler state as parameters and keeps the normal resolver-information constructor.

These are medians from the existing repository benchmarks, with at least 50 rounds on the same CPU.
The async-list and pickle-decoding slowdowns are smaller than the larger interquartile spread from their respective runs.

Related: #142 discusses async execution overhead; this proposal measures synchronous introspection.

Environment: CPython 3.12.13, Linux x86-64, AMD EPYC 9R14.
Base commit: f9bdf7c5c728a04bd3f27bfb8d394e9ac95751e2.

Reproduce on the base and changed versions:

PYTHONPATH=src python -m pytest tests/benchmarks/test_introspection_from_schema.py tests/benchmarks/test_async_iterable.py tests/benchmarks/test_serialization.py::test_pickle_large_query_decode --benchmark-enable --benchmark-min-rounds=50

The full test suite passes: 3,613 passed, 18 skipped.
Ruff, formatting, and mypy also pass.
Codex helped prepare and test the change; candidate results were tracked with Weco Observe.

Would this small Python execution change fit the project's maintenance goals?