Ok, so after reading Jack's response I feel like backing my claim up a little better (i.e. doing a little self-dissection).
I'm not a physicist, but I do study energetics of natural ecosystems and human societies, so I'm familiar enough with energy to explain it a little more thoroughly than I did before. What we're interested in is potential energy, because the potential energy stored in the limbs of a bow will end up as the kinetic energy embodied in the arrow. There are several types of potential energy, including gravitational, electrical, chemical, and elastic. We're obviously interested in elastic potential energy. Elastic potential energy is calculated:
E = (λ χ2) /2 l
Where:
λ is the elastic material's modulus of elasticity,
χ is the elastic material's extension beyond resting, and
l is the elastic material's natural length or length when at a resting state
The bow's natural length represents the horizontal length of the bow from the point on the string where the arrow is nocked to the front of the bow's shooting shelf. This is, effectively, the bow's brace height. The extension beyond resting is the distance from the string at the arrow's nock when the bow is drawn to where that point sits in space when the bow is braced but not drawn. These points equate to the braced bow's setup envisioned as if it were a spring, with the braced bow being the resting state.
Note here that the only variables that influence the potential energy at full draw are these two distances. The wood's modulus of elasticity is constant (or at lease we're assuming it to be). So this means that the potential energy generated by drawing the bow will be the same in both bows, and again since most of the arrow's energy will be given to it very soon after release the fact that the bow's tension at brace is different doesn't matter. It doesn't factor into the equation.
Alternatively, one might argue that the bow's resting state is actually as an unbraced bow. In this case there would be a difference in potential energy stored if the deflexed bow's tips were not reflexed enough to being them to the same plane as a bow without deflex. I'm not convinced that this is the right way to look at the problem, though, as the websites I looked at to put this together all showed a braced bow when describing its resting state. Just for argument's sake...
Overall, though, I'll stick to my guns. There's no reason why a deflexed bow with lower tension when braced won't transfer just as much kinetic energy to an arrow as a bow that isn't deflexed, given the same limb, string and arrow mass, brace heights and draw lengths.
All the best,
-Eric Garza