@J-Hizzle this might a good read: https://link.springer.com/chapter/10.1007/978-3-031-97570-7_4 we could try to compare MRMD against those numbers if we ever do a JOSS paper or similar.