Originally Posted by
WineCountryUA
Interesting to see all the other schemes tested, although I didn't intend the above to be an absolute statement about most optimal (it does wind up reading that way I suppose).
Seems in these evaluations they don't test front to back though, which should be slower than back to front.