If we could use GPT-4 to grade prompts, we wouldn't need to talking about grading prompts to use for GPT-4, since this solution requires that the problem doesn't exist.

You can't use the thing you're testing to evaluate its own performance. This applies to rulers, speedometers, and AI. It's the difference between a "subjective" and "objective" metrics. If you want an objective metric, you need to have it based on something external, based on reality. Otherwise, you have ideas holding themselves up: nonsense.