I would only consider this as “in depth” if and only if the author provided clearer methodology on the exact model used for this write up.

It is still unclear to me whether this is a deficiency in all ChatGPT models or this is just one of the many.