I read extensively about other systems of variable cam timing last night. There is not very much information on the tuning of such systems, particularly with a turbocharger.
There are some SAE papers that are written to optimize cam timing around NOx and fuel economy, but not around torque or spooling the turbo.
From our friends in the Toyota, Honda, Mazda, and Mitsubishi communities it looks like the intake cam is advanced beginning around 3000 RPM and continues up to around 6000 RPM in the vacuum regions of the their VVT maps. The tuning of these regions was determined empirically on a dyno and no amount of uninformed futzing was going to see much of a gain. There is a theoretical discussion of timing of the cam to match the resonance of sound pressue waves on a Mazda page, but that would seem to have more to do with their variable intake path than with actual cam timing given a set intake path as in the Subaru.
Advancing the intake cam in the early boost areas seemed to be the more prevalent mode for turbo applications, and in this case the more advance possible without a loss of power/torque the better. All the Subaru and Mitsubishi maps I've seen would seem to bear this out. None the less some information on the torque produced in said regions would seem to be the ultimate guide to tuning of the intake cam advance.
Here is some good data and a half decent discussion from a Mazda page:
I also looked into when in the 4 stroke process the injectors begin to fire. This would get at the question of whether or not fuel is being blown out of the induction charge and into the exhaust. I could not find any specific information on our motors, but in general the injectors do seem to begin firing BTDC such that fuel will be pooling behind the intake valve prior to it's opening. There is a great deal of gas velocity/inertia physics in here that would determine if any fuel is escaping out the exhaust valves. I'll have to talk with some other more technical folks about our motors and how much of an issue this may or may not be for AVCS intake cam tuning.
For now, everything points to the need for dyno to determine an AVCS map accurately. The the impediment of exhaust flow through a larger turbo or through a more convoluted header path probably affects the map somewhat, but the intake manifold and cams would seem to have a larger effect. I think if I were to try and optimize tuning of AVCS it would be around MAP vs RPM vs Torque just like any other fuel or ignition timing tuning that I might do.