It could be, as already suggested the wastegate/vane actuation of a VGT, though I think it's a wastegate on these.
I'm unfamiliar with this models exact turbo set up, but I presume the wastegate/vane is actuated via an arm into the turbo which is operated via a vacuum or electronic box.
If it's visable, it should be wastegate open (arm at one end of it's travel) with the engine off.
Start the engine (idle) and it should move to the other end of it's travel and wastegate is closed.
Increase RPM slowly towards the red line and it should move back towards wastegate open, this is how it limits boost as rpm/exhaust pressures rise. (if it doesn't, it's likely to detect overboost and trip limp, hence why it often trips when acelerating)
If it's not performing like that, either there's a vacuum leak on the pipe, a leak in the diaphram of the actuator, (or electronic box is knackered) or more likely the wastegate in the turbo is coked up and stuck.
This last one can be checked fairly easily, if follow the actuator arm to the clip and remove it, the arm into the turbo is free to move back and forth.
Work it like this, if it's stiff, it's coked up and a bit more working is needed to get you by (until it cokes up again) or you clean the turbo out.
There have also been a few duff reading from MAP sensors reported, these read the air pressure in the air intake after the turbo.
A faulty reading should be easily spotted if you have a diagnostic tool to read it's output and could over read and tell the ECU there's more air under pressure than there actually is, though the coked up wastegate/actuator is more likely.