The modulator should hold vacuum and not leak, but if it's a slow bleed off, that shouldn't affect shifting drastically. The old type modulator has a rubber cap, the new type has a plastic cap with an O-ring. If the rubber cap is cracked, or O-ring isn't sealing on the plastic cap, this could account for a slow loss of vacuum. What you describe sounds like a moderate leak if it takes 10 seconds to leak to zero (double-check all the fittings, including at the Mity-Vac itself).

With this cracked line fixed, your shifting should improve significantly. Next step would be to replace the modulator itself so you have a modulator that holds vacuum properly, then set the VCV per FSM, and tweak the modulator slightly (if needed). I would also try (for grins) bypassing the BFS entirely and connecting the modulator directly to the output of the VCV. Oh, and yes, you can't test the VCV with a Mity-Vac, since it's a controlled bleed... you have to test that with a constant vac source, but your previous testing pretty much proved that it's working normally (12" at idle, zero at WOT, variable in between).

The Bowden cable only affects part-throttle shift RPM. If you're happy with the shift points, leave it alone.