@aggarg, I have had several cases where I have a task that takes “commands” in a queue, and at times that task needs to queue up other operations for it to perform when it finishes what it is doing right at the moment. Rather than create a new internal method of remembering that need, it will just post that operation to its queue.
It can also make sense to use if you need a quick to design implementation of a small stack or fifo, since the code is already written, you are just paying a bit of overhead for the things you aren’t using. For that case, you likely always use a 0 wait time.