Unless the users are not getting imported for some reason (they already exist?) then the number it reports is the number importered.
Ah, but if you are running it after the initial import then that number is inflated because it’s counting users that existed when the script started. So that’s about a week to import those.
How many posts?
How fast is the machine? How much ram? What have you done to tune postgres? Is your source database a bottleneck?
I’m not aware of a secret bulk import script for smf2. I have had an import that took weeks to run. The final import won’t take that long. I’d just stick it out.