The more journaling mailboxes you utilize, the more you can take advantage of Archive Attender multithreading. It is hard capped at one thread per mailbox though. In order to ensure archiving keeps up in the event of a failure (especially if you are using a single journal mailbox), a good idea is to have tiered archiving policies as sort of layered message traps by date:
An example of the policy to be on the front line would be:
Other policies would be between dates, with the final policy applying to anything older than the second to last policy's criteria.
I've noticed significant improvement once policies were added. This method also sort of safe guards you against back logs by already have a system in place to deal with them.
If you are bent on journaling at the Hub Transport level, you may want to at least split internal and external mail into two different mailboxes. If you are journaling at the database level, then you could go as far as to have a separate journal mailbox for each database.