I still have not been able to completely eliminate the stutter. I have noticed that my stutter is significantly worse when the weather is cold, but the engine is fully warmed up. (cold weather + warm engine = stutter).
By playing with the load compensation table, I was able to get it so my total AF corrections were always within +/- 5%, but the car would still stutter. This got me thinking that if the issue is really some sort of resonance in the fueling or intake system, that it could very well be only affecting 1 or 2 of the cylinders. Since the car only sees an "averaged" AFR from all 4 cylinders, it seems reasonable to me that even when the AFR's are spot on, one or two of the cylinders could be rich or lean.
This led me to digging around some more in the map and I found the per cylinder IPW compensation tables. There is one table for each cylinder. These tables have some pretty extreme values of correction up around 11% in places. They also have two very obvious peaks around 1600 and 3000 rpm. Those happen to be about the same rpm where my stutter is.
Out of curiosity I decided to test what these tables do. I drove the exact same route twice. Once was with the stock IPW tables and stock load compensation table. The second time I completely zeroed all 4 of the IPW tables but again left the load compensation table alone.
I then plotted total AF correction vs rpm and got some interesting results. I have a attached a picture of the graphs side by side. The left is the stock tables, the right is the zeroed tables.
It is obvious on the stock tables that there is a peak around 1800rpm and a valley around 2800rpm. With the tables zeroed, the peak at 1800rpm remains relatively unchanged, but the peak at 2800 rpm is eliminated or even reversed.
I am not sure how to interpret these results yet, but those IPW tables obviously have an affect on AFR's and can be tuned per cylinder.