when the message is domainkeys identified mail and at the same time it is spam, it looks like that checkpoint is not able to flag subject nor header. The message is properly identified as spam (I see it in log) but the subject is not changed thus we cannot use next logic in following mail server which depends on this.
I tested it by sending the message with "RPD Spam test: Spam" in subject from several free emails. All messages were properly detected as spam but only messages from email servers which haven't used domainkeys were properly marked.
We have version R75.30.
Can anybody else test this behaviour, before I send ticket to checkpoint