I have not seen this discussed, but I completed an Osage bow a couple years ago and it has some character to the limbs. The lower limb had a deflex in the last 6 inches and the top limb had a reflex in the last 6 inches. I tillered it as it was and the limbs appeared even when strung and drawn....life was good.
Reading all the heat bending posts I decided to make the lower limb reflex like the upper limb, so I used my new heat gun and put in the reflex in the last 6 inches. Now both limbs appeared even when un-strung. The crazy part was now when strung the bow does not match on both ends. The lower limb appears straighter and different than the upper limb. I did not remove any wood from either end.
I know I can re-tiller the limbs to match, BUT here is the question that I have; Given the wood is the same. I wonder if I had "not" truly tillered the limbs to match to begin with, I just tillered them to "look right" to the eye.
If a bow was securely fastened into a vise and a scale was hooked to the knock end and then pulled 10 inches, should the opposite limb have the same pull when tested to the same specs.
We measure a bows pull weight by measuring the pull on both limbs. One could be 26 and the other 34 equaling a 60 pound bow. Perhaps by testing each limb to the same pull is more important than seeing tiller that is pleasing to the eye. I know the lower limb is suppose to be slightly stronger and by testing the limbs individually this would be more accurate.
This will perhaps apply more to "character bows" than to board cut designs of near perfectly matching limbs.
Some of you may already test each limb separately beyond floor tillering. If so please share your method. If you have a character bow or even a non-character bow try testing the limbs separate with equal bend to see what weight they pull.
This was difficult for me to explain in writing, so I hope it makes sense.