That actually wasn't it.... and I'm not sure what did it in the end. The index wasn't present, so I ended up creating the index, only to have the upgrade fail again. I restored the DB yet again, verified there was no index on the restored DB and ran the upgrade one more time before I was going to contact support.... and that time it ran to the end without issue. I have various log files showing the problem, and then it just ran through. I don't get it at all, but I guess that doesn't matter, as the service is upgraded and running.
My guess is that in the original work last Friday, something had made the installer choose the wrong DB upgrade script, and when I ran it yesterday and it was successful, the correct script was chosen.
I say this because of the log entries from the start of the first run with the upgrade failure. My DB version was absolutely v4.1, and it ran the 5.0 script.
[6/21/2013 3:50:58 PM] Info: Upgrading database from version 5.0 to version 5.1
[6/21/2013 3:50:58 PM] Upgrade-vMN-to-vMN-Next\MSSQL\upgrade_mssql.sql
Then this from the first failed run yesterday where I had erroneously put the index in place in attempt to fix it:
[6/26/2013 9:41:41 AM] Info: Upgrading database from version 4.1 to version 5.0
[6/26/2013 9:41:41 AM] Upgrade-v410-to-vMN\MSSQL\upgrade_mssql.sql
I don't have the log from the success, as it cleaned it up. But something was wacked out on 6/21.