How do Twitter screen names and user ids actually work for feeds?

I am having trouble with inconsistency making API and RSS calls to Twitter status feeds. For API access it seems that for some Twitter accounts the screen name is sufficient, but often it's important to get the proper upper casing (e.g. DylanValade vs dylanvalade). However开发者_StackOverflow社区, on other accounts the API sometimes fails for the screen name but works if I hunt down the numeric user_id and use that instead. Here are three URI's that all do the same thing and return either RSS or JSON but when I call them programmatically I can get null returned in some cases and I can't identify the pattern.

http://twitter.com/statuses/user_timeline/dylanvalade.json

http://api.twitter.com/1/statuses/user_timeline.rss?screen_name=DylanValade

http://twitter.com/statuses/user_timeline/134889007.rss

Does anyone know if or how Twitter treats their screen names separately from their user ids and which service is the preferred approach?

Found a good note here finally in a non technical article about RSS dying

Looks like Twitter RSS requires the user's numeric id while JSON can be requested via user name.