Sounds to me like it's an issue with their VLM captions creating very "pretty" but not actually useful captions. Like one of the example image prompts includes

Sounds to me like it's an issue with their VLM captions creating very "pretty" but not actually useful captions. Like one of the example image prompts includes this absolute garbage: > Convey compassion and altruism through scene details.