The existing options --{read,write}-{rate,burst} are per connection.
The new options --worker-{read,write}-{rate,burst} are per worker
thread, which is overall rate limit of all connections worker handles.

The profiler and benchmarking showed that calling evbuffer_add()
repeatedly is very costly. To avoid this, we buffer up small writes
into one large chunk and call evbuffer_add() less times.

some small additions for clarity. please check to ensure meaning is preserved.