The scope of the experiment was, to see if the splitting surface holds up as a back. Even mostly unscraped and with imperfections.
My thinking is, the back would have failed before the belly, when overstressed, if the imperfections would have done too much harm. And that is supported by what you say: the wood is always more tension strong, but the sides would have changed, if using the raw splitting surface, changes the tensile strenght properties.
But exactly that isn‘t the case, the belly failed and which shows, that the back is still stronger than the belly.
In short: I wasn’t about to show if vertical grain is stronger. I guess you haven’t read all of my posts.
@brad:
Do you mean the one I showed in my last post or when I redo the experiment?