Hi all, first post and a little long winded so bear with me. Just trying to be thorough.

I need some help with the diagnosis of my intermittent turbo boost. This car is mostly driven by my daughter so I'm not behind the wheel often enough to correctly log the symptoms prior to any attempted repairs. I noticed the check engine light on and scanned the OBD2 code which came up with P0299 turbo underboost.
I cleared the code and within a day it was back. My daughter did complain about the performance of the car but whenever I drove it it seemed fine, not realizing it was an intermittent problem. Talked with a guy I use who does work on my cars and he said the turbo's probably failing and should be replaced. We ordered a new aftermarket Turbo, manifold and wastegate combo and he changed it for me and gave me back the car saying it's running great. He did say, and showed me, the original Turbo he removed had cracks in the exhaust manifold (which after some research, seems to be common for the OEM Land Rover Turbo from Borg Warner.)
Thinking of the problem was solved, it threw up the underboost code again within a day. Now I figured it's time to pay attention and log the symptoms using my odb2 scanner and reading the data stream in real time for MAP (manifold absolute pressure.)
For a reference point, my scanner shows the ambient barometric pressure at around 30" hg.
What I have observed consistently is when starting the car cold and getting on it I can achieve up to 56" hg (12 psi) MAP with excellent performance. After driving anywhere for 5 minutes or so I can no longer achieve anything higher than 33 MAP under light to medium acceleration and on flooring the gas the map actually drops down to about 26 indicating a vacuum in the intake manifold. Of course there's no boost, power or acceleration under this condition and will continue to be this way for as long as I have the engine running. Once I turn the car off and let it sit for about 5 minutes and restart it I will have again 56 on the MAP pressure under hard acceleration and the Boost is only available for about 3 to 4 minutes before it again caps out at 33 under moderate acceleration and again dips into the twenties on hard acceleration. I have logged some max boost pressure in the 40's before it dissappears entirely so it seems to be a fading event rather than abrupt loss of boost. So it's not a question of a cold Engine/turbo versus a warm engine/turbo but rather it seems turning the engine off for a few minutes allows a reset condition of sorts that allows for the boost pressure to be available again for a short time. Idling the car for 15 minutes does not reset the boost availability. The engine must be turned off in order for the reset to occur.
The original Turbo fan spun freely when removed from the car with no noise roughness or any other indication of malfunction. Other than the cracks in the manifold I believe there was nothing wrong with the original Turbo and the problem is elswhere. I don't think it's pressure leaks as that would be constant boost problems instead of predictable intermittent behavior.
Any ideas?
Thanks