There are some TGV deletes on the market that have been known to be faulty, for instance if it was replaced with a cheaper composite TGV delete the actual composite housing could be cracked. Or there could be a gasket failure in any number of locations after the throttle body that would cause this issue. Your uncle's computer will always read that their is a TGV sensor code. If I hooked my Snap On Modis up to your car it would read the samething. But what it seems like is this car has an Opensource tune, and the previous owner went in with ECUFlash and just disabled the codes from triggering the CEL to come on for that particular parameter. Even though any scan tool will still read it as a faulty code. As [MENTION=9]HolyCrapItsFast[/MENTION] mentioned, with something that has modifications to this level, it's important to know what's going on with the engine as a whole. We can help you diagnose this problem in further detail with data logging the car.
And I do also agree with [MENTION=662]Batmobile_Engage[/MENTION], this does seem to be more along the lines of a mechanical issue. A leak like he said will cause a loss of power, especially when boost pressures enter the manifold and pressurize it. If there are failing gaskets that would be a tall tale sign for where your problem may here in lie.
One thing you should really consider doing is a Boost Leak Test. This test will pressurize the intake system and allow you to check for leaks anywhere with in the intake track.
A scan tool, is not the end all do all tool. There have been plenty of times where a scan tool has led me down a rabbit hole, and if good old fashion exploratory investigation happened, a major frustrating headache could have been avoided. For instance, had a car come in with a misfire code, as the scanner said. Pulled the plug, plug looked fine. Started looking at the coil pack, checked resistance to the coil. Coil was fine. Looked at the rubber boot on the bottom of the coil, and it had a small pin hole in it, was leaking electricity through the hole in the boot and grounding out on the spark plug well. Replaced the boot and the car ran fine. But had I just went with what the scan tool said, it would have cost the customer $205.00 for a coil pack, instead of $5.00 for a new coil pack boot.
A little investigation and leg work, like data logging or a boost leak test could save you hundreds in parts and or labor charges.
I hope this helps you out. If you need any further help, don't hesitate to ask. We will all try to pool our combined knowledge to give you a hand, right boys? [MENTION=1172]Alin[/MENTION] [MENTION=9]HolyCrapItsFast[/MENTION] [MENTION=662]Batmobile_Engage[/MENTION]