Makes perfect sense. Out of curiosity, I know scryptN is usually referred to as "about half the hash as scrypt" but is the 0.47 a proven value?
So the issue there is the 0.47 changes per GPU/CPU. Some cards will be 0.40, some will be 0.55, varying by clock speed, memory available, memory speed, etc. Generally higher memory cards (or faster memory) will have a slightly better edge on Scrypt-N vs Scrypt. 0.47 is just a pretty good estimate for scaling (I think this is based on R280x cards - a pretty standard mining card).
When we're stable with the additional algorithms (X11/X13), I'll be adding some custom stuff to allow you to switch your miner dynamically for the most profitable algorithm (I think, it will need a TON of testing - don't hold me to it). If that's the case, it will need to have a more accurate measure for each miner's speed change.
Long story short, 0.47 is pretty widely accepted for the hashrate change (seen 0.45, 0.47, and 0.50 in different places).