feat(tooling,fw): improve type checking dev experience for pydantic models, e.g. `Storage`

This would reduce friction for test implementers unfamiliar with the EEST codebase.

Here's an example: If a `Storage` object is defined in test cases, `mypy` fails to detect that `int` is a valid type as a dict key or value. Currently, such `Storage` definitions need to be either marked as `type:ingore` or expliclitly cast to `HashInt`, e.g. `Storage(cast(Storage.StorageDictType, {1,1}))`.

Here, for example, type checking is manually disabled inline:
https://github.com/ethereum/execution-spec-tests/blob/2e1d3a0c7ec7a2f02fd7b9ad4900f79e323910f1/tests/cancun/eip4788_beacon_root/spec.py#L61

Otherwise, `mypy` returns the following two errors:
```
tests/cancun/eip4788_beacon_root/spec.py:61: error: Key expression in dictionary comprehension has incompatible type "int"; expected type "HashInt" [misc]
tests/cancun/eip4788_beacon_root/spec.py:61: error: Value expression in dictionary comprehension has incompatible type "int"; expected type "HashInt" [misc]
```