Yes, that is correct.
'RMS' or 'Stable Averaging' is where you average a point in the last spectrum measured, with the same point in the current spectrum measured. This is the usual type of averaging. The average spectrum is then returned.
'Peak Hold' averaging isn't averaging at all really. What you do here is compare a point in the last spectrum measured with the same point in the current spectrum measured keeping the highest value. The resulting spectrum can therefore have a mixture of last and current data.
'Exponential' Averaging averages a point in the last spectrum measured with the same point in the current spectrum measured but uses a weighting factor to bias the average towards the current spectrum. So if you had a transient event say it will eventually die away as more averages are taken.
Regards
Steve