It is no coincidence that what is in the charcoal and the quench effects the colors. You can carburize steel in pure carbon, and vary the final color exactingly by temp at quench....
If what you say is true about differing temperatures being able to create different colors, do we need polymorphs. I suspect quenching to create colors is actually poor heat treating practice. I think the dunking is intended to be a random or partially random interrupted quench. Maybe, that's how the same piece can have surface effects that're typically created at different temps, maybe.
I don't think you have to carburize steel to show different colors. I think you can take a torch to just about any polished steel and create temper colors, iron oxide in varying layer thickness.
The other thing I'm not quite grasping is the thought that you're heat treating iron oxide in a similar way that steel is heat treated for hardness. I think iron oxide can exist in atypical forms, but aren't they very to extremely rare? As mentioned a bit ago, are iron oxides taking an atypical polymorphic crystalline form because of stored energy from a reaction, or was it just stacked up that way by a form that's since been removed?
Edit to add that I can appreciate that you're discussing how short lived some of these structures may be.
