It appears that the first pod status update tick is fired on average 1/2 pod status frequency duration. This is unnecessarily long - I don't know if we have an existing issue to cover this, but it makes many small, simple pods take up to a minute to report ready (on their first time). @vmarmol added a note to convert this to a better mechanism, but I wanted to make it obvious.

Some thoughts:

• The first transition between Pending and Running is the one that should fire with minimal latency - we should look to send these ASAP, in preference over all other status updates
• Subsequent updates can come at lower priority
• We should probably poll/check/update pod status at different intervals (inversely proportional to their age up to a cap), since most failures will happen in the first seconds of launch