Brad, I don't construct them or tiller them in order to baby the patch. I want the patch to work like the rest of the limb. The bow pictured above is a fairly highly stressed semi-recurve design. 60" ntn and 65# @ 29". The outer limbs were kerfed with an additional osage lam slid into the recurved section, which if anything, adds even a bit more stress to the mid and inner limb areas of this bow. The patch is located approximately 3-4" outside the dips... it's working pretty dang hard... and now dyed and sprayed, you'd have to be told it was there and then hunt hard to find it. It looks as solid and integral now as it did 7 years ago.
Del, I follow your diagrams. I just don't fully agree with them. I didn't comment on your cherry pit comment because I think it's not directly relevant to our discussion, and I was trying to be brief. I understand where you're going with it, but I think the relationship of these patches to the bow limb are more complex than you believe or show in your drawings. The one I showed a picture of was ground through any compression forces, through the neutral plane and right up to the backing strip, at least along the outer edge... yet it wasn't ground across the entire width of the limb. It's ground into the limb at an angle. Why grind across the entire width of the belly if you don't have to? ... and grinding them at an angle, too, helps incorporate them. Another positive aspect of this type of repair.
Since the day this patch was tillered into this bow, it has acted homogenously with the limb. There have been thousands of arrows shot from it, and it practically looks like the day I sprayed it... other than the dye has faded a bit.
Again, if properly mated, prepped, glued in, and tillered, the joint will be strong, quite strong, which means the glue joint isn't an issue.... and so long as the glue joint isn't an issue, the 'outside' wood trying to 'spit the pit', can't.