I mean any sort of critical thinking, numerical reasoning, deductive reasoning, error checking tests etc are pretty objective.
If I were running a recruitment process for example, I'd have an application form with 0 questions.
Automatically allocated test (based on grade requirements) with WG+ deductive reasoning + numerical reasoning.
Automatic timed long answer question test based on passing benchmark - I'd have 4 questions randomly allocated from a question bank of 40 questions (10 questions per type). I'd publish all questions online so that nobody would have an unfair advantage from communicating with people who took the test.
Then AC.
Of course, no process can be completely objective, but I still think SJTs are a non-evidence based way to thin out applicants. (Of course, you could argue that people who practice and research SJTs deserve to do better on average, and to an extent I agree)