I disagree. Their spam filtering has been top notch for me. Sometimes spam does get through, but marking that as spam has made sure the next ones of the same are flagged correctly.

Lately LLM generated spam has fooled it more often, though. Let's see if that becomes a problem.