Are those 4" tiles? That would give you maybe 2" of change from unbraced to braced? I was messing with lots of deflex and I started to wonder if there was a "law of diminishing returns" going on. I mean if you have to reflex the tips and limbs to make up for the draw weight that the deflex stole is there any gains to be made? I guess you get draw length on a short bow, anything else?
I am getting about three inches of tip movement at brace. Doing it that way, based on a typical 6 inch brace height, I am actually able to use half the stored energy that a 6 inch brace has. So even though string tension is lower I am able to get more energy from a low brace. This design works best with full contact recurves because as you draw back the bow gets longer and helps to negate the diminishing returns you correctly mention. They are still there but just reduced. Another thing is, because of the deflex, the string has a leverage advantage over the limb tips, this requires the limbs to be more massive to fight the leverage advantage. I feel with the more mass in the limbs, there is less set because you have more wood fighting the stress. I could be way off base there though but these short bows have some very thick limbs. I will post some dimensions later today to illustrate my point.
I will add with reduced string tension the bows are much quieter on release and with shorter limbs, tree stands and tight shooting quarters are much more doable. I want to make a compound bow sized self bow that is a viable hunting bow and standard yardage. I feel I am damn close to it.
Point is, when you build this bow you are instantly at a disadvantage on energy storage. Everything you do after is to reduce the effects of lower energy storage, and use the energy you have more efficiently. The recurves, wider limbs, reflexed limbs, anything you can do to increase string tension and reduce tip mass and keep stability. So far, it seems if its done right a very effective bow can be built. My arrows on all these short bows have more than enough penetration as demonstrated at OJAM with some of our fellow forum members.