I think what the OP will find with some testing and/or experience is that his theories are correct, it's just a matter of applying them as needed. I have a bunch of experience mounting gates on gasser stuff. Mounting the gate 90 degrees to the path of flow will not use the exhaust pressure to help open the gate as it would if the path of exhaust pushed directly on the head of the gate. Diesels are of much lower rpm and speed of exhaust flow, so my feeling is the manifold becomes more of a slow moving, pressurized chamber of drive pressure, and having a larger gate does get the job done. Also, you might be surprised how little the valve opens to achieve the desired pressure reduction. Therefore, having the inlet pipe crushed has less effect than the 90 degree mounting.
You wouldn't be the first to successfully weld the gate inlet pipe directly on the outside of the turbine housing. Laying it over so that it has the flow going directly into the gate will help it work tremendously. The trouble is keeping it from cracking. I have done a preheat and tig with 309l rod, on cast iron, and daily driven my truck about 2 years without any failures yet. I have not used this weld only to support the weight of a 50 mm gate against the vibration of my Cummins, though.
I would be more concerned about dumping the excess back pressure with the nitrous setup than using the gate to drop drive pressure between compounds. I think the back pressure of the atmosphere turbo on the outlet side of the gate slows the flow down a lot vs. dumping to atmosphere, and actual flow becomes much less, and you're really just balancing pressure in two chambers, so to speak. In this example, the 90 degree mounting works just fine, on my truck at least. I can tune the back pressure right out of the manifold, keeping it 1:1 up to around 70 lbs. of boost, then it starts climbing.
I have pics in my phone of the gates on a 6.4 Ford I'm doing for a guy, let me see if I can get them to show here....