It is about High Availability. Have you ever tried to use a central broker in NSB? The best HA you get is clustering 2 machines. And that is not HA.
For HA you simply run a distributor process for each logical event on windows cluster and add as many worker nodes as you see fit.
Because NServiceBus uses the store and forward pattern if a NServiceBus process/machine hosting it goes down, you are still guaranteed the eventual delivery of messages when the process/machine is resumed.