The limits of analogy are important, but in this specific context so is the full breadth of definitions. I concentrate here on “flagellum as motor.”
Etymologically, Latin “motor” means the one who moves something. Usually a person.
It’s fifty years since I did physics A-level, and though I still get regular dreams about being unprepared for the exam (never for zoology, for some reason) I did revise that in mechanics, a machine is defined as “a system using energy to perform work.”
Some of the simplest machines are inclined planes, levers and pulleys, and the laws of mechanics apply however the system is set up. Thus an inclined plane used to build the Great Pyramid is no more or less a machine than wind blowing sand up a sand-dune, and a falling tree dislodging a boulder is as truly a lever as Archimedes moving the earth with one. Similarly, the olecranon process of the elbow is a lever, even if it also performs other functions. No analogy, just application of mechanical definition.
“Motors” are the class of machines in which the form of work produced from energy is (primarily) motion. So they may be electric rotational motors, linear motors, clockwork mottors or rocket motors (with, at their simplest, no moving parts). They are motors if energy->system->kinetic work.
And that is why, when I studied physiology and anatomy, striated muscle consisted, functionally, of motor fibres, the muscle was supplied by motor neurones (hence “motor-neurone disease”), and those ended up in the motor cortex. No analogy - just a biological instantiation of mechanics.
It matters not that muscles are also used to generate heat (in shivering), sound (as in laryngeal folds) or whatever - any more than a siren ceases to be a motor because it has some extra device, or because it is also a gyroscope.
The only question, then, is whether the flagellar assembly converts energy to motion (of the flagellum, or of the cell). If it does, it is a motor. And it does, literally and not analogically. The comparison MUST be with the definition in mechanics, and not with some popular image of what a motor “ought” to be.
By all means, at that point question any claims to irreducible complexity, to design or whatever errors you believe the IDists make. But if you’re going to use Shannon Information theory, rather than a broader definition, to discount biological information, then to be consistent you should use the definition of “machine” or “motor” derived from mechanics to assess molecular machines or flagellar motors. I don’t see why there should be any dispute on those definitions.