Comment 2 for bug 1582720

I was looking at this one today during the Hack-A-Way some. We can certainly go over the name value pairs and prevent duplicates from going in when they're processed in misc_util but I'm increasingly of the thought we shouldn't. There's a question about to the extent that code should compensate for bad data and if the label, note and href are all duplicates I think that's bad data and something that catalogers should be alerted to to clean out, not cleaned up by the system.