Earlier quoted context omitted.
FWIW, every server-level spam filter I'm used has had so many false positives it's caused me significant problems. Also FWIW, people to whom I send email regularly tell me that their email systems mis-classify my emails as spam. I've spoken with several networking specialists, and they all shrug their shoulders and say: "It's the server level spam filtering ... it's not very good." GMail in particular is known to be…
Just as a counterpoint, my company uses Google Workspace for email and we receive a good deal of spam. Been using it a couple years now and it's been very good at identifying spam, with very few false positives.
I have/work with three different companies ... two of them use Google based email systems, one (my own) does the filtering "in house". One of the G-based systems performs well, the other has multiple false positives every day.
I'm sure it performs well for the majority of people ... it's likely that the emails I send and receive are atypical, and Google performs badly because they are not "normal".
Everyone has to decide where that trade-off lies. In my context, I simply cannot trust Google, or any other server-side filtering, for important email. As others have said, it simply means the spam folder is a second inbox.
As I say to my clients about so many things: take advice, understand the trade-offs, make a decision, deal with the consequences.