I generally make my bows with a longer top limb, and preordain every nock point 3/8" above perpendicular to the shelf.
I then tiller for limb balance while drawing the string where my string hand fulcrum will be. When it's built balanced relative to that nock point location and my fulcrums, there's no need to move nock points afterwards to 'fix' anything. There's nothing to fix.
Tiller profile at brace is merely a result after tillering is finished... not a beacon to guide me during the work. It simply is whatever it ends up as. Nowadays I don't even bother to take that measurement because it's meaningless. Sometimes it's negative, sometimes even, sometimes positive. It can vary from bow to bow, even if they're layed out and shot the same. How it ends up depends as much on the unbraced profile, any inherent internal differences in the limbs, etc. That can all change from bow to bow, as can their tiller profiles, while each is equally, accurately balanced.
Since my tree mimics how I shoot, I know if it's balanced on the tree, It'll be balanced in my hand and shoot an arrow just right, first time, every time. It's so simple and predictable.
Conversely, if I had two bows, the 1st with a little deflex in the bottom limb, where the 2nd bow was straight or had reflex, and tillered them both to the same predetermined tiller measurement... cuz that was my standard practice... which would be balanced/correct? One? Both? Neither? How would I know? Cuz the arrow flew poorly? Then what, move nock points to try to correct imbalances? Sounds like some unnecessary guessing and hoping to me... then fixing problems I built in. No thanks.