So the base engine (aside from mechanical strength!) was and is never the issue, its the bolt ons and adaption of water injection or aftercooling spraying to keep charge temperature under control and so long as you have the appropriate turbo its a prefect linear relationship.
It is and can be swayed by these variables:
* If you are greatly reducing ign timing as boost increase
* If you are making vast changes to AFR as boost increase
Then you can have non linear discrepancies of over 25%!!!
But if you run water injected especially these boundary restrictions do not apply
To give you a practical proven example a BMW F1 turbo mapped on an engine dyno here (I have the dyno sheet!) is Australia did 802bhp @ 3.3bar with lambda of 0.79 @ 10,000rpm and at 5.6bar uses the same lambda and water spraying the IC it see's over 1350bhp That was the going rate at the time for the qualifying power output at that boost pressure