I have mutated my setup even further. Forever tweaking...I guess
While my original idea for this thread worked, what it really did was get me thinking outside of using a single DAW, and the limitations to make it work for every studio situation. Needs for tracking, VST instrumentation, and mixing/mastering, are all different. I either end up with multiple version of a project, and constantly tweaking my DAW, or I find the least common denominator...which tends to not be the best solution for each situation.
AS of now, I am still running eMotion LV1 V11 to control my Soundgrid setup. I actually prefer it to SoundGrid Studio V11.
I have found the StudioRack Native to be just as useful as StudioRack SG, with less annoying undocumented features for using it with Cubase. The part the doesnt work well for SGS v11, is the I/O naming. Constant struggle with naming changes in my templates. Very frustrating.
I also like running SG host on a dedicated computer, separate from my DAWs. No need for that being overhead on my DAWs.
StudioRack v11 latency, even with a SG extreme server, measures out to be about the same as local host for latency. At least in my experience, you dont need a SGS to take advantage of the lower latency. Maybe you do if you have a wimpy DAW, but since I am running on a OC 7980XE, my native mode latency is just as good as the SGS. I also have the ability to route things out to LV1 or SuperRack and back into the DAW if i wanted, with a similar RT latency.
As I said earlier, one of the major changes I made to my workflow is making better use of two DAW's.
I decided to dedicate a recording DAW just for live studio recording/tracking. I set it to 64 samples, and that puts my RT latency to about 5ms. That is useful when I am playing back guide tracks for the artists, while tracking.
My recording DAW is a 5960x OC at 3.8GHz. Plenty of power for studio recording. I dont really even need any effects on this machine, as the studio musicians are using the effects in LV1. Simple, super clean, and very low latency.
For studio HP monitoring, I am using LV1/MyMon. It works fantastically. I run 3 cue channels from my recording DAW, right into LV1, for click and any guide tracks needed for the artist. By using MyMon, and LV1, the musicians basically have no latency, and they can do their own cue HP mix with their phones.
When the tracking is done, I can edit/comp tracks to my hearts content. It can be a giant mess, and I don't care. I dont have to hide it, or make a new project to put the comp'd tracks into. I can save it as is, in case I need to come back to it to add or redo a track. When the comp'd tracks are finished, I just use Cubase feature "import tracks from project" from my recording DAW, into my main DAW, over a shared network project drive.
Since I am not using my Main DAW to record live performances, I can do all my VST tracks and effects for the project without so much concern about absolute low latency. I am fine with 128-256 samples for VST work, and even higher when it comes to mixing, since latency is far less important at that point.
On my main DAW, I am still flipping back on forth between VEP, Cubase disabled tracks, and Cubase track presets for VST instruments. I guess I am still trying to find my "flow". VEP is becoming less and less. I suspect I will eventually move away from it for good. It is liberating to have a clean project template, and then just load up the instruments as I use them via disabled tracks or presets. These are some of the great features I love about using Cubase.