Yes, hitting API rate limits correctly returns 400 error codes as
documented. I am not speaking about the api rate limit. I am talking
about a limit to rate that status updates are allowed for a single
user. Sending status updates does not count toward a user's rate limit
(as verified by monitoring rate_limit_status while sending updates).
The "tweet limit" I speak of occurs after sending updates with high
enough frequency. When some threshold is reached, new updates sent to
Twitter are not accepted. There is no error returned when this limit
is reached, rather the update_status method simply returns the last
previously successful status update for the user.
Since no error is produced nor is this published in the documentation,
I would simply like to have a maximum target rate so that I can
throttle updates to stay within the limit.