I haven't followed or read this entire thread, so sorry if this has been discussed. Without the VNT lever being allowed to pull the VNT control arm to the stop you are giving up the key spooling/tip-in region of the mechanism's dynamic control range. It seems to me this is greatly negating the greatest benefit of your turbo choice.
Have you looked into shortening the VNT control arm like we have to do with the GTB turbos?
Fix Until Broke said:
A5INKY - You're absolutely correct, however this turbo is small enough that even with the vanes wide open (no actuator at all, it spools almost identically to a well tunes 17/22. The issue is a geometry related, not stroke length related. The actuator is barely perpendicular to the arm when it's wide open and flattens out the more you close the vanes so it gets to the point where it just can't move any farther. Posts 2 & 3 show all this so you don't have to read the whole thread.
So, what A5INKY said has been rolling around in the back of my mind for a while - more on that in a minute...
Jeff and I worked on the tune at TDIFest and in short we made the N75 map "steeper" so it opens up the vanes at higher RPM/Fueling more than it did before. The goal was to reduce EMP's and get more RPM out of it without compromising transient response. I had the actuator adjusted to start closing the vanes at ~7inHg.
We were generally successful at lowering EMP's without losing any bottom end response, however the more I drove it the more I realized that we made the right changes, but it probably needed some more fine tuning as the smoke on tip in was worse than before (13:1 was pretty easy to achieve). Not having the knowledge or tools to adjust the tune myself I figured that adjusting the actuator would be equivalent to moving that whole map up/down and I could at least experiment a bit. I couldn't change the slope of it or any individual points/areas, but I could offset it.
Now, back to A5INKY's comment. I decided to make a big change in the actuator adjustment and shortened it as much as it would go. I was able to get the actuator arm really close to the stop screw (0.25mm/0.010") at ~20inHg and the actuator moved with any vacuum applied.
Took it for a test drive and EMP's flew to 50+ PSI in 1st gear at relatively low RPM's (2000 ish), however in 2nd, 3rd, 4th etc gears the EMP's spiked to ~45 psi upon spool up (1400 RPM), boost to ~20psi, 20:1 AFR and then the N75 went to zero, EMP's came down to match boost and it reved out to ~3500 RPM before the EMP's got back up to 50 PSI.
As I type this out I realize that I should probably have a graph or at least a video of my dash to go along with this - we'll get that done.
Long and short of this is that I think A5INKY was right, the actuator travel was not enough to fully stroke the vanes. By shortening the actuator a lot (5+ turns), I was able to significantly improve the transient response of the system through significantly increased EMP's during spool up. This makes sense once you think about it - more differential pressure across the turbine, the faster it will accelerate. This also reduced smoke since there was more air sooner and, since we steepened the N75 map as mentioned above, the vanes were essentially wide open* once the turbo spooled up which reduced EMP's
*This turbo will spool on it's own at higher RPM's, however at lower RPM's (below 2000) with the vanes wide open it won't spool, or will be really slow to spool. By adjusting the actuator really short, when there's no N75 command the vanes are somewhat shut. So with the setup I have, when I adjust the actuator it ends up adjusting either or both the beginning and end of vane travel. As I adjusted the actuator longer and longer, the spool up didn't change much because there is still no pre-load, but the 2000 RPM EMP's, Boost and AFR's all dropped with each adjustment longer of the actuator. Since the actuator adjustment moves the whole "map" and only a portion of the vane travel for this turbo, it's a compromise.
Most of this rambling is me keeping notes of what I've done and what happened. The other takeaways from this are that it takes a lot of work to properly setup/tune a turbo to perform it's best where you want it to (always a compromise) and you can significantly improve transient response by increasing EMP's during spool up, but it's best done in the mapping. The increased EMP's don't seem to negatively effect performance, or at least the gains from spool up and airflow outweigh the negatives of increased EMP's for a net advantage.
I've thought about putting a valvespring in mine originally, but a friend had a bunch of spring sets laying around from his high dollar automobilieo. 
