This avoids the send_self_retract hack.

This ensures that retractions are sent too.

This avoids receiving stale data in answer to the update.

Now that requests work well, there's no need to increase seqno_interval
as fast, and this makes it more likely that routes will be feasible.

Updates being delayed for longer than 5 seqnos may be legitimate.

This avoids repeatedly sending requests to a source that has just rebooted
and lost its sequence number.