The retry mechanism has been removed in the latest version of the module (the mechanism behind scheduling queued actions has been reworked).
The retry mechnism unfortunately is a necesity. When you schedule a Process in the queue the data that is being created might not be available in the database yet. If you would immediately start the parallel action it could be that not all data is present yet, we haven't been able to find a more reliable way to determine when a microlfow process has finished.
In other words disabling the retry will likely break the logic.
Can you elaborate why you would want to skip this retry mechanism? Or do you just want to disable the FIFO to speed up processing?