I think the inductor one only makes sense if you imagine the flywheel being driven with huge buckets -- such that the voltage drop is massive if ∂V/∂t is large

I think the inductor one only makes sense if you imagine the flywheel being driven with huge buckets -- such that the voltage drop is massive if ∂V/∂t is large (with analogy to L ∂I/∂t). I'm far from convinced that is totally right though, as you can easily have arbitrarily high voltages with switched circuits and inductors, but not necessarily with turbines... I guess a transistor is a little man twiddling a valve according to a dial in the "water model" of electricity, any through-space effects are due to leaks, and nonlinear components like MOSFETs are a bit more creatively explained. (E.g. Source: water reservoir; drain: water reservoir; gate: gate between source and drain reservoirs, driven by an utterly mad person who follows interesting rules. Oh, and inexplicably he's a bit stronger than transistor man too).