CentOS just did a whole new batch of updates, and something broke sed in some circumstances. So, I’ve rolled a new version of the install script that replaces the affected use of sed with a perl command that does the same thing, which is not impacted by the bug.
That said, there is still one more place (that I know of) in the install process that is affected by the bug, and I won’t be able to fix it until tomorrow. It exhibits as a failure in the Postfix configuration step. You can workaround it by rebooting your server immediately after installation completes, and running the following command:
# virtualmin config-system --include Postfix
Then proceed to logging into Webmin and doing the post-installation wizard, etc.
This will be fixed after I get some sleep.