Not exactly what I was going for. Rather, it's half of what I'm trying to accomplish. Your suggestion is essentially what I ended up doing... the only solution I could come up with. Technically, it will work, but it's not very satisfactory due to the fact that the half's counterpart has to be created by the same method. In other words, take the bottom loft and stitch to copies of the inside and outside patches (or create two new ones altogether using the loft edge as you did... which IS better). But, as I understand it, stitching doesn't fill gaps with patches, it cinches up two edges. That's why the lines being stitched are required to be the same size (citation: help docs). When the lines are cinched, they are usually altered a bit. So, when you try to introduce a third line into the same stitch, it pukes. Leastwise, that's my best guess.
If that is in deed the case, then this method will produce two halves that are not a perfect match. The mating surfaces may have gaps or worse, intersections. Neither situation is desirable in a mold. The result is inexact in a world (of CAD) where precision is typically carried out 8 or more decimal places.
Nevertheless, I ran the part on the mill this morning, using this very method. I won't know the exact variance between the two parting lines for a couple hours. But, if you think about it, you'll realize that the parting line, in my two half mold (A/B) is comprised of two nearly identical parting lines. In my way of thinking, there simply HAS to be a way to create a single parting line that can be used to create both halves. If the variance is less than .0001" then no problem, but stitching the surfaces in my drawings have resulted in tolerances in excess of .1" If stitching shifts my edges even .005" and both sides of the mold do that in the opposite direction, the gaps they leave will be unacceptable.
Anyway, thanks for trying!
Regards.