i have wondered about this too. the neutral plane argument seems reasonable enough, but that would mean that the mini bow isnt adding anything that a thicker bow, or some other backing, wouldnt add. the design is certainly harder than a cable backing, so why not just use a cable backing?
this is just a thought, i have not made nor have i studied any of these bows myself. however, i was thinking that based on microscopic reversibility, the micro steps you take to get from A to B are the opposite of the steps from B to A-what goes up comes back down in a symmetrical fashion. The mini-bow flattens out against the big bow early in the draw, and at that point it can add weight but it seems liek it would be stacking weight-in other words it shouldnt add much in energy storage. the big bow doesnt start to bend at all from the tips until relatively late in teh draw. this means that the bow tips of the big bow would stop moving early in the release as well. as the opposites all happen, the tips would hit home well before the string is finished moving forward thrusting the arrow. the little bow now starts to pull off the big bow and snaps the string taut using the long bow tips as leverage. it would be the energy storage benefits of a short bow with a longbows mechanical advantage-same idea as a holmegaard. the energy curve would be steep early in the draw as you pull the short bow into stacking mode but that would be translated over a longbows length, so late in the release you should have the same steepness of energy released. SOM