This is exactly why I am trying to determine where Porsche hobbled the engine to keep the 911 on top.
If I understand the question, the answer is: through programming of the ECU.
Consider: so-called Stage 0 or 'flash only" tunes increase power in turbo engines without any hardware upgrades, solely by altering code or maps (table of numeric values) in the ECU. If they can increase boost and power that way, then Porsche can limit boost and power in the factory programming. So factory-rated peak power does not represent the full thermodynamic potential of the engine with all systems (fueling, boost, ignition, cooling) running wide-open, but an arbitrary limit imposed by how the ECU marshals it all.
Actually I have this vague idea that the factory-programmed "limit" is not just a particular number for max. boost, but a more complex set of maps which basically describe a predefined power curve. So on a dyno the engine is just drawing the power curve with which it has been mapped (or the one which best fits the exact current conditions of fuel octane, internal temps, ambient air pressure and temps, etc.). The automotive version of ascendancy of computers (software) over hardware. Could this also help explain why dyno results and factory bench-test ratings don't seem to align with real-world performance for these cars....is the computer actually recognizing when it is being tested and selecting a soft (sandbag) mapping?*
I think tuners in the past mainly increased boost by manipulating Wastegate Dump Cycle WGDC code. With VTG, the same actuator which controls vane position also controls the integrated wastegate. So tuners may be encountering a learning curve in this area with the 9A2/VTG engines (2.5 and 3.8 911 Turbo/S/GT2).
*I think this forum has still not answered "how is this even possible": that a nominally 350hp 718 accelerates to a range of speeds/distances at exactly the same rate as a 400hp 991.1 Carrera S with the same PDK and weight? (per C&D track sheets, quarter-mile and all splits below 120mph)