Make validators consistent

[brianhelba]

• Use Iterable for parameters, since the input is never sliced as a sequence
• Use list for attributes, since it's always actually stored as that type
• Use consistent import of _ValidatorCallable

AI Policy

• I have read and agree to the AI Policy, removed any "Co-Authored-By" lines attributing coding agents, and manually reviewed the final result

[github-actions]

Diff from mypy_primer, showing the effect of this PR on type check results on a corpus of open source code:

zulip (https://github.com/zulip/zulip)
- zilencer/migrations/0026_auditlog_models_extra_data_json.py:17: error: Module has no attribute "serializers"  [attr-defined]
+ zilencer/migrations/0026_auditlog_models_extra_data_json.py:17: error: Module has no attribute "core"  [attr-defined]
- zilencer/migrations/0026_auditlog_models_extra_data_json.py:24: error: Module has no attribute "serializers"  [attr-defined]
+ zilencer/migrations/0026_auditlog_models_extra_data_json.py:24: error: Module has no attribute "core"  [attr-defined]
- zerver/migrations/0452_realmauditlog_extra_data_json.py:17: error: Module has no attribute "serializers"  [attr-defined]
+ zerver/migrations/0452_realmauditlog_extra_data_json.py:17: error: Module has no attribute "core"  [attr-defined]

[sobolevn]

Thanks!