the simplest way of thinking of this is that when a bow is drawn, it stores a given amount of energy, which is not the peak draw weight. two #55 bows can have a different amount of stored energy. the efficiency is the amount of that stored energy that is imparted to the arrow. once again, different bows will have different efficiency. many things will effect that efficiency, string type, silencers, serving weight, as well as limb and tip design. OL has come up with one way of getting a high efficiency and as far as norb's tests goes, the blackswan was as high if not a miniscule amount higher. sooooooo, in general a recurve will have a higher stored energy than a longbow but generally a lower efficiency and hybrids blur the ground in between. and yes, the difference b/w most high performing bows at 28"amo and 9gpp is about 4-5 fps and the same gpp but 30" amo is about 5-6 fps which is all OL has ever claimed, which is equal to about 1-2" of draw length which means in general, an acs cx or other very high performing bow will allow an archer with a 26" draw to shoot like he had a 28" draw. if that means alot to you, then great, if not.......fine, wouldn't argue about that point of view either. the 20 fps difference many talk about is only b/w the absolute top performing bows and the absolute slowest out there.
once again, if you see tests out there that show it to be a greater than that they're generally highly flawed, and the only way to know is to scientifically test it, feel is subjective and a bows properties on draw and on release an affect how fast an arrow seems to be going so relying on feel or sight is very very misleading in general.