old p-40b,c,e,m are mess!
The irony is that you're actually right in stating that the "flight model is a mess"... from a certain point of view. It all depends on what criteria are used to evaluate that statement.
In an
ideal case, a simulation of a physical thing should work like real world when you insert real life parametres into it.
But since simulators are always approximations, sometimes you need to fib things a little bit. Add some correction factors somewhere, adjust another properties elsewhere, and then the simulation
looks like it's working right... which is actually the only thing that really matters to
end user gameplay experience.
Looking at things from physics perspective, it actually
does look awfully fishy that the aircraft requires such a high engine power value to achieve its performance specification.
For example, if you look at a simple physics simulation where a box with 100 kg weight is lifted to 1 metre of elevation in one second, that requires exactly the power of 981 W.
If the weight of the box is changed to, say, 1000 kg, then to achieve the same elevation in same time, the power must also be increased by 10 to 9800 W.
Now, the mass of the box could be set to any figure and there could always be a power level that results in same lift time, and if you make a video of it, it should look exactly the same in all cases.
But even though the simulation gives the same "lift time" performance in both cases and will look similar in that respect, the higher mass and higher power variants will have other differences in a more complex simulation. Increased inertia and moment of inertia come to mind first. If the object is lifted by a cable or rod, the tension or compression forces are much higher, leading to different amount of deformation, different resonant frequencies, different structural strength requirements etc.
And this is just a simple idealized situation where a device of some vague description lifts a weight to a prescribed elevation in a certain amount of time.
Increased amount of parametres leads to exponential rise in crosstalk effects between each of them. Adjusting one parametre may result in unwanted changes in the behaviour of the simulation elsewhere. Something like a flight model has a lot of parametres, and it is sometimes difficult to adjust them all so that you can insert "real world" values into something like engine power, and actually get correct visual results out of the simulation.
A proper evaluation of
why the power needs to be set so high might be interesting, but it would require immense amount of testing and changing FM parametres one by one to see what effects each of them have.
Again from physics perspective, if an aircraft in a simulation requires too high power to reach its prescribed top speed, then it would suggest that the aircraft's drag parametres are set too high (which in itself could be caused by high drag coefficient, low lift coefficient, or high mass, to name a few possible reasons) or that the propeller's thrust efficiency ratio is too low (specifically at high speeds at the altitude you're testing at). That's just a conclusion from the fact that, at top speed (zero acceleration), the thrust and drag forces must be balanced and equal.
But unless you're interested in doing this as an academic exercise or self-flagellation of some kind, there is very little to be gained by actually doing it because the "fixed" flight model should retain almost exactly same flight characteristics and performance as the current flight model.
I would rather see modders and Team Daidalos use their talents for improving the game in a way that is apparent to us players.