One of the experiments, in progress. There have been past discussions on the use of needlewheel pumps to serve a dual purpose. Though, not a whole lot of follow-up with empirical data that I could find. Data forthcoming.
Ttystikk noted waterfalls which have been reported as one of the most efficient mechanisms. Flooming as an option as well. One benefit include not needing a separate air pump(s) and stones, among others.
Do you feel air pumps/stones are necessary or, more specifically, could be something that might be used as a point of comparison? Rate of O2 uptake higher than generally thought? Hint, hint: collecting compelling reasons to run an experiment