The whole issue is very obviously LLM generated nonsense. The stats are way too specific and reinforce the user’ bias in typical hallucinated fashion.