It does seem to raise fair questions about either the utility of these tools, or adoption inertia. If not even OpenAI feels compelled to integrate this kind of

It does seem to raise fair questions about either the utility of these tools, or adoption inertia. If not even OpenAI feels compelled to integrate this kind of model-check into their pipeline, what's that say about the business world at-large? Is it that it's too onerous to set up, is it that it's too hard to get only true-positive corrections, is it that it's too low value for the effort?