If someone sends me spam, with a from/return say gmail address, and my spam software attempts to return it after rejecting it - will gmail not see my server as spammer it self? I ask because it seems my IP is getting limited for high levels of unsolicited mail, I can see what else would be triggering this flag.
using exim/cpanel
What you are dealing with is "backscatter", and your mail server's IP address will certainly end up on several major blacklists if it is configured to indiscriminately send bounce messages (non-delivery reports) to the return path address of an undelivered message.
Ideally, if your mail server is not going to be able to deliver a message for some reason, it should reject it during the SMTP transaction. That way the sender's mail server has a record that the message could not be delivered, and the onus is on the sender's mail server to inform the sender that their message was rejected. If your mail server accepts a message and then for some reason cannot deliver it, the best course of action is to silently delete it (black hole it).
A side note (and this is only one of several additional considerations) is that one of the reasons to reject messages during the SMTP transaction is if the recipient's address doesn't exist on your mail server. However, doing this will allow a "directory harvest" attack, in which a spammer is able compile a list of the valid addresses in your mail system by brute-force (as your mail server helpfully points out which are valid and which aren't). One way to mitigate this is with tarpitting.
Your question is tagged with spamassassin, but you don't mention what mail server you're using. If you update your question with details of your server set up, you might get an answer with some information about how to configure your mail server to avoid this problem.
A good place to start to get an understanding of this is the Wikipedia article on backscatter.
