Einstein was reduced to doing lecture tours for money in the 1950s (not unlike Tony Blair), and to make him more comfortable, they gave him a mature German-speaking chauffeur, who used to sit at the back of the lectures rather than in the cold car.

One day, the chauffeur joked that he could probably present the lecture as well as Einstein, and Albert suggested he did so that evening. The chauffeur presented the lecture with slides and notes, very successfully, while Einstein sat at the back of the hall wearing the chauffeur's cap.

At the end of the lecture, a member of the audience asked a penetrating and incisive question. Without missing a beat, the chauffeur replied: "Your question is so trivial, I will call on my chauffeur to answer it!"

Over to you, Jon!

My first instinctive response to this was to refer to Lee Smolin "The Trouble With Physics", which relates that James Clerk Maxwell realised that electrical and magnetic fields generate each other in a way that propagates radiation in a straight line. From this theory he calculated the speed of the radiation and found it was identical to the known speed of light. Great unification! Speed of light determined by solid maths, so c is a universal constant.

At 04:00 today, woke up in a hot sweat. I would be OK if c was a dimensionless constant. But it is 2.98E8 metres/sec. Metres are 1.0E-8 of a quadrant of the earth's circumference along a line of longitude. Seconds are 1/86400 of the period of the earth's spin (relative to its varying position in orbit, no less). [Yes, there are now 'invariant' standards for these, but these are essentially copies of the original arbitrary units.]

So how does c have the gall to define itself in metres and seconds? What if mankind had standardised on the well-know furlong/firkin/fortnight measurement system instead?

My conclusion is that Maxwell's calculation has its roots in some experimental determination of the electrical and magnetic field strengths, which then introduce some combination of the units those experiments measured.

So the "universal constant" c is dependent on the assumption that electrostatic and magnetic forces have universally invariant constants of proportionality in them somewhere. Considering we do not understand gravity as applied to galaxies, and have had to invent unobservable dark matter and dark energy to retain universal gravity, I expect it is only a a matter of time before somebody needs to invent dark magnetism and dark charge in order to prop up the myth of invariance of c.