It's a common misconception that marking a message as 'not spam' causes the sending address to be whitelisted. It doesn't. It merely flags to plusnet that a message has been falsely identified as spam, so that any general issues can be taken up with Postini.
Messages sent to your plusnet address(es) go first to Postini for spam filtering. If they come from a plusnet relay, they routinely get identified as spam (I think it has to do with the sheer volume of traffic Postini sees from this source). This is why the domain 'plus.com' is whitelisted for us all at the organisation level (and so the most blatant spam spoofed from our own addresses gets past Edge Protection and has to be stopped by plusnet using a relay check). However, if I've got this right, hosted domains send via the plusnet relays without the org whitelisting. To avoid the spam filter, I think the sending address will need to be whitelisted by you, the user.
Hope this helps.
