Distinguishing spam from google servers

I run a very good anti-spam setup, except for one nagging little problem: Google.

Some legitimate users use Google's mail service with custom mail domains. In other words, they do not use a [email protected] sending identity, but their own.

I find that almost all SMTP traffic from a mail*.google.com host which is not from a @gmail.com sender, however, is spam. There is not a lot of it, but it's a nuisance.

These mails with custom domains don't seem to have any clue as to the Google identity of the originating user, so it is impossible to report the user to Google. You also obviously can't just blacklist Google servers because it's a very popular e-mail service with a large number of legitimate users.

(Of course, a [email protected] sender could also spam, but that doesn't seem to be a big problem at all; I can't remember the last time I saw a spam that directly identified a gmail account in any way. That makes me unconcerned about @gmail.com senders.)

Is there some good way to block the spam (ideally reject it at the SMTP level, without even looking at headers), while avoiding false positives: blocking non-spamming owners of custom mail domains who use Google for sending mail?