Despite all your calculations you do forget several factors.
I didn't calculate anything.
The water does not scale down.
The efficiency of model props is notoriously bad.
I am well aware of the so called
scale effects.
But because model tests can only be performed at Froude's scaling law, and not at Reynolds' scaling law (to mention just the most important two) at the same time and scale (other than 1:1)
there are introduced a whole bunch of correction factors and correlation coefficients during the test procedures that I left under the carpet not to totally confuse readers here.
BUT
Despite this there is one over riding factor.
Actual testing.
I only was referring to actual testing, and this is nothing that I made up.
If you really are interested in the model testing procedures and the corrections that are applied for the inevitable negligence of scaling laws other than Froude's
then I recommend you have a glance at the
1978 ITTC Power Prediction Method which was a recommended procedure for all towing tanks and ship model testing facilities world wide to be followed to arrive at comparable data and predictions based on that.
You may find a list of all recommended procedures issued by the ITTC
here.
I also fond a more tangebile outline of the model testing procedure in e.g.
this thesis.
Yes I am familiar with Froude as an ex Chief Engineer so can categorically say you are not only wrong, but so wrong to be embarrassing.
Beg your pardon, I cannot see what I should feel embarrassed about.