Honest answer: We really #$%^ed up the aggregator yesterday by throwing approximately a half million errors while moving every database for all sites in the network, really, all of them.
The aggregator takes events that happen on any site and pipes them up to a central database where your account list is pulled from, pretty much everything on your SE.com profile comes from there as well.
We knew this huge move was coming, we knew it would cause issues, we knew the data would be FUBAR afterwards...that's why this issue hasn't been fixed in recent weeks. To fix the data Monty has a backfill process to re-populate these network tables, similar to when the "global" database was first created...but we have to take API down for a bit to do it in any reasonable amount of time. Because of that we don't want to run a backfill, then another, then another to fix the issues from recent weeks due to infrastructure moves.
That being said, the largest move of them all happened yesterday and we're keeping an eye on any further issues. We'll try and get a backfill run ASAP to sync up numbers everywhere.
I really do apologize for all the errors yesterday. I tried to keep as many as possible from leaking out to the world while keeping the sites up (even if read-only sometimes) as much as possible during the moves, only taking maintenance when we really had to. It was a very complicated situation that we'll get a post-mortem up on next week. Everything went fairly well with the database moves themselves, it was at the end that things got very ill a few times. That and some lessons learned for anyone else making similar moves are things we like to share, because we love you.
All the guys who try and keep this running