well, perhaps all the reflex happened before you started working on it, or maybe further reflexing was somehow prevented by the compaction of the belly that happened during tillering or from the resistance of the normal wood in the bow to bending. I am not an authority on wood or bows, by any means, but with my latest bow project, I have been playing with some reaction wood, albeit compression wood, and it acts different for sure, especially when adjacent to normal wood in the stave. I did not write my suspicions clearly in the previous post, but if the bow took a remarkable lack of set when tillering, perhaps the expected set was hidden somehow by tension stresses in reaction wood. I find it interesting that your description of the bow in its present state, seems similar to the high early draw weight one would expect to find in a highly reflexed bow. maybe that is why Badger suggests that the weight per/inch draw increase will drop offas you near full draw.
By the way, do you happen to know which way the back of the bow was oriented in the tree? Was the pith of the limb off center?
Did you know that Native Americans often preferred branch bows because they had not yet met Dan Perry , and even if they had, it was too long a ride to find a thickness sander.