> A COMPLETELY DIFFERENT THEORY.
Rust is iron oxide.
There's like 16 different iron oxides.
He shows conversion of Red (Fe2O3) to Black (FeO). The red oxide happens with ample oxygen, forms hard, coarse, and flaky. The black oxide happens when oxygen is available but limited, forms as microscopic balls which make a sticky "mud" which is easily crushed.
The scratch on my car fender gets an infinite supply of oxygen, is forming red oxide, our familiar "rust".
My steam boiler and radiators make lots of black oxide. There's liquid water to bring oxygen and iron together, but after the first firing there is not a lot of loose oxygen in the system. Each cycle, a bit of air gets in, all that damp iron sucks it up so fast there is not enough to go around. I tap the mud and it is pure black. I let splash dry in the drain and it brushes right off.
A radiator which has been abandoned, open to damp air, forms red rust inside. This may be black oxide converted to red oxide by ample oxygen and some damp to bring the O and Fe together. When I put a rusty radiator in service, I must drain a lot of black mud for the first month. I dunno if the red oxide goes back to black in the low-O system, or if the heat breaks red rust off and exposes raw iron which forms black oxide.
If you keep the system near-sealed, the black oxide covers the surface and oxidation nearly stops. If you leave the system open for a summer you get LOTS of black mud next fall.
So the oxide conversion makes sense. I never knew you could force it in minutes with soda and voltage.
The material of the counter-electrode is probably immaterial. Since we don't need great conductivity, and the process may attack the counter-electrode, and scrap steel is cheap, that's probably best.
There are rust treatments which do form black oxide with chemistry. The brush-on glop I have used (not Naval Jelly) was weak at forming black oxide, and was cut with other stuff to bind that black oxide to the surface. If it worked better it would be good for hidden areas of cars (and is widely used for that). The stuff I had was about as effective as a coat of good enamel, but messier.