Interesting thought on the vacuum breaking the active siphon, but I am betting it would only break if the momentum of the water slowed to a point at which it could not force its way up and over. It is not like the vacuum gets in the way of water flow, nor water on the down-side of the vacuum bubble. I am betting there is a mathematical way of figuring this critical water speed out but I am not a fluids guy. Necking the pipe down on the upper elevations may be a way to speed the water to a point it shoots up and over.
I would be willing to bet that if you put a pump on the high-water side, you would only need it to start the process, and then valve-in a passive bypass once going. Would a hand pump be enough? The act of pumping obviously takes air pressure out of the equation as you force water beyond the 34-foot limit.
