I see I forgot to adress this issue...

As Billy allready says, each engine has its individual "set of defects", but it is not only that: in the total chain of control mechanisms, there are also parts that are OUTSIDE the engine.
For example, the linkage ratio between throttle servo and carburettor throttle lever is never exactly identical from set-up to set-up.
there are differences in the fuel tank set-up, fuel level height difference, different props, different air-resistance and flying speed, propsize etc etc.

So much so, that even if I would take the engine off from plane A, and fit it to plane B, even if I would also transfer the throttle servo and the valve and its electronics, then the linkage between servo and carb alone would force me to redo the curve. Not by much, perhaps, but it will NOT be "spot-on".
It is just the nature of this system.

Therefore, it IS possible to run, test and set up the curve with the engine in a testbench, but it is NOT possible to then move the engine etc into the plane and expect the curve as found on the bench, to function in the air. It would no doubt be close, it would probably even be a useful "starting point" but it won't be the same.
This means that if you want to experiment with different props, indeed you will need to as a minimum thoroughy test and check the curve, and in case necessary modify.

In general, the shape of the curve will be the same, but it needs to be tweaked.

But two identical engines of the exact same type, on two identical planes with identical props still can require pretty different curves because of the accumulative influences of these individual differences.