I wouldn't consider that a definitive test either. Whether or not it changes color with heat will have more to do with the type of dye the anodizer used then it will the Type classification.
There are a lot of things that determine the type. Thickness, Weight, Process, Type of Acid, Rate of Growth, Current Density, Bath Temperature. The thickness is, however, one of the few observable/measurable qualities, and as such, is usually the confirming test for Type III.
Anodize of any type should not "rub off" very easily. You have to remove the aluminum oxide in order to remove the anodize - which is amazingly difficult to do.
What do you think could cause an anodized coating to fade with only exposure to skin oils and a couple showers? I have one which has lost quite a bit of its color (click for bigger versions):
This shade of turquoise is a relatively subtle color though, so look what happens if I change its hue by 60 degrees to show how the same fading would look in blue: (all I changed was hue, nothing else)
Was this even anodized at all? Could it have been anodized and dyed but not sealed? Any other ideas what would explain this?
Does this mean we can have white coloured anodized type 2 lights?
If this is possible, I want to request that every light from now on is white anodizing.
Anodizing dyes are technically not opaque - they are translucent. Because of this, white is pretty much impossible - it will always be tinted toward the color of the base material - in this case, grey.
Yep. One other nice feature of the coating is that it will survive to whatever temperature the aluminum itself will survive. In other words, the coating will not be damaged, degraded, discolored or removed by heat until the point at which the aluminum melts.
The process of applying CeraFuse involves plasma discharge with temperature over 5000 Celsius… I wonder how much it costs to apply this to lights. How practical is this?