Certain principles that exist which make it hard to "cheat" the systems. Stuff exactly like how a bow with lots of reflex should shoot faster, but reflex CAN cause extra set, instability, higher limb mass, and are harder to tiller properly. Net gains may or may not be realized. Likewise, recurves easily store more energy, but often have issues with tip weight and stability, thus net gains may or may not be realized. As such, IMO, it comes down to how much you push the envelope, and whether your skills are up to the task.
I have sketched out some RIDICULOUS designs I will probably never get around to trying, hoping to "cheat" these principles. It would be great to achieve radical profiles that still manage to relieve strain on wood, increase early draw weight, counteract oscillation/vibration, dampen lateral wobbles, stiffen tips, create zero stack, etc... Stuff like hollow bows, Superlight limb tips, weightless recurves, two stage recurves, string bridges, inverted limbs (incredible F/D curve, but totally stupid in every other practical way), cable backs, string stoppers, bizarre limb and handle arrangements, etc..
What WORKS for me tends to be; enough mass in the right places, MODERATE reflex or recurve, skinny tips, heat treating, Perry reflex, good workmanship, and knowing how much is too much to ask of your materials.
To the original post: Baker covered a lot of this in the first TBB Design and Performance chapter, though sometimes he mentioned it in passing and didn't dwell on it. Less set or even reflex increase performance. Deflexed tips help bows survive to higher draw weights. How do thickness, width, and length affect stiffness, set, and mass? He stated that less or more of a limb's bending potential can be used up by high brace heights, short working sections, reflex or recurves, etc. So, yeah, any given limb can only bend so far, and if much of that potential is used up before we get a string on it, less is left to get it to full draw. It's a balancing act.
I don't know how you add up the numbers exactly, though. I don't think you can add one atop the other in a meaningful way. I think you have what you have once you start pulling the braced string, and high early draw weight is good. You just have to get there