That sounds much better and is around the norm under modest ramping, makes much more sense.
The hill scenario you describe represents a far higher "ramp" and it is definitely possible to see the boost your referencing to under such conditions, i use a "control" ramp to compare like to like setups which is the same as a flat road rolling on in fourth from 1500rpm which is less than the hill type scenario but almost identical to the on road conditions. We simulate the hill type ramps during tuning also but they give unrealistic boost response times.
Please dont think im criticising your setup as it sounds awesome, hence my reference to "artificial ramp" in my first post, everyone checks the boost response different and is why it is only useful to use the same dyno and tuner for back to back testing, or a set routine or road to road test the changes..