Last night (31 July 2013) at 10:30 pm PST I saw a spike of 500s when collecting users. There were some 404s in the mix too, which was a little concerning. It’s the second time I’ve seen this in recent weeks.
The endpoint used was https://api.twitter.com/1.1/users/show.json passing the user id as the id param.
These user ids were among those that had 404s:
There were also 2 calls to a user’s timeline that received 404s. Once at 10:30 amongst these other errors, and once at 11:20pm. The user’s timeline url used was https://api.twitter.com/1.1/statuses/user_timeline.json with user_id and since_id provided, and include_entities set to true.