Well, as I think I understand it...... At say 68", one of my forms took the limbs back 2" at mid-limb and then forward 4" from there, so the tips ended up a couple inches ahead of the handle to start. R/D is already pointed back at the handle area, say the first 25% of the limb, and this, coupled with the fact that a tiny amount of bend here makes a big movement at the tips, means it only needs to bend a little here. Most of the bend is in the middle 50% of each limb.
So, if the inner limb deflexes sharply and the tips sweep forward gradually, the outer 1/2 of the limb is straighter. I it's almost straight to start with, any tiny bend you allow it can straighten it out completely. This takes less advantage of the string angle and braced tensions design features that make the R/D design so nice. You are almost forced to bend the inner 25% more, the next 25% bend more, bend the outer portion not at all, or it will straighten out all the way. Not the end of the world, but if you have some curve there, the inner 25% can bend a little, the middle 50% a lot, and you can have a stiff tip and a nice reflex hook at the tip. If you don't the inner 25% has to bend more, the next 25% more, and the whole rest of the limb less, or the limb will go totally straight, or even start to bend toward the tips, too much.
Hope that made sense at all. The "low-stack" design in the TBB by Adam Karpowitz is fine, too, and nothing wrong with a setback handle. When I made this mistake trying to R/D a laminated mollie way more than I should have, it worked OK, because i was planning on the long stiff lever and it stayed reflexed because it was stiff. I guess the takeaway is that less bend in the outer forces you into longer and longer stiff tips.
