That's assuming the boat isn't already moving.
The analog to the water's drag would be resistance in the coil, I presume. This is why I say we should be exploring means of minimizing the coil's resistance (such as coating the wire with graphene, which has the lowest room-temperature resistance of any known material, prior to insulating the wire). It'd be akin to putting that boat up on hydrofoils.
Now, if the boat were already moving (even slowly), it'll take less to keep it moving. After all, the pistons and drivetrain in that boat's engine are nothing more than a percussion translated into rotation translated into linear motion... they're the "hammers" hitting the stern of the boat via the prop.
Once the initial inertia is overcome to get either the boat moving or the coil "spinning", it takes less to keep it going. The lower the resistance, the less it takes to keep it going. Reference MRI magnets, which emanate a magnetic field with very little power input, because they're cooled below the critical temperature and hence are superconductors. Some gravimeter superconducting magnets have been
running for as long as 22 years with no additional power input.
The only other way to do it is to exploit the induced magnetic field and applied frequency to generate phonons which will cause boson pairing (Cooper pairs)... in essence converting the electrons into a room-temperature bosonic superfluid, which exhibits zero resistance. But the engineering to do that would be far beyond the capabilities of... well, everyone, everywhere, thus far. Except, apparently, Newman himself. And possibly Searl.
One problem I foresee is that phonon frequency in copper is
highly temperature-dependent... so we'd need some sort of temperature feedback loop which adjusts frequency to maintain phonon density.
Another problem I foresee is that the Debye temperature (corresponding to the phonon cut-off frequency) of copper is 158.63 F. Go above that, and you get no phonons.
A third problem is that
the phonon frequency of copper is quite high... approximately 6.6 - 7.3 THz. So the coil would have to be designed to 'ring-down' near this frequency. Conversely, iron has a phonon frequency of ~5.39 - 10.49 GHz... perhaps that's why some experimenters used iron wire. Which experimenter was it who said they'd made more power from iron than they ever made from copper?
And of course, the main consideration as regards phonons is phonon density, which is dependent upon the differential energy variation within the coil... so we'd have to ensure the coil has sufficient energy gradient per unit time to maintain phonon density... meaning high voltages and quick on/off times.