Hi, since SPOOL stands for "Simultaneous Peripheral Operations On-Line" it implies that jobs have to be buffered (in memory or on disk) first to be able to be executed correctly. Usually it's associated with IO from/to devices because these things operate a lot slower of course than the 'executioner', i.e print spooling (on disk) in a printer queue.
I would take 'output spooling' as a general form of buffering (in memory), which also can take place between processes, like in this series of commands
cat textfile | sort
where the output of 'cat' has to be buffered first before 'sort' can operate on it's data.