Forgot to post back here on Friday with my findings!

The root of the problems stems back to the 25th June when an internal problem (ref: 57120) was raised to deal with the fact that catch-all customers were receiving separate Quarantine notifications for each alias a spam message had been sent to (instead of one notification containing a list of all the aliases that had received a spam message).
This problem occurred shortly after some work was completed where we had to fail over to the backup Quarantine server. After comparing the configuration of the two servers we noticed that the backup server wasn't quite set up the same as the primary. We corrected these discrepancies and for a week or so we thought all was well.
That was until we realised that the notification emails were getting seriously delayed. We raised this with IronPort who requested access to our network so that they could do some digging themselves. After chasing them for a response a few times they came back to us and advised that they believed the problem to be the result of a known bug. To fix this it was recommended that we try upgrading the software running on each of the servers. This was done last week (28th) at which point we could see that notifications were being sent from the boxes again.
Sadly this success was short lived and the sending of the notifications stalled again. We went back to IronPort who got in touch on Friday to advise that they'd changed some variables in the back-end systems and that things should now be working again.
If people still think there are problems then please shout up.