Not true, otherwise scale model testing of large ships would not work. The reason scale models are used in tow test tanks is because the effects can be corrected by using the Reynolds Numbers to adjust the effects.
So you'd have to use something other than water, with a very different Reynolds number (I wonder what you'd have to use to simulate something that small?) which means that again our intuition about what would happen in water is irrelevant.
Also as you point out that's used for
tow tests. No-one tries to simulate the actual mechanism of propulsion, which would be more difficult. The non-scaling relationships between length and area and volume and mass would get you every time. But even if it could be done, then, as I say, our intuition, based on water, doesn't tell us how the experiment should turn out. There's no use for common sense applied to conditions with which we are unfamiliar.