A multitude of things from the specific engine condition and operating environment all the way to improper ICP calibration or within the repeatability of the ASTM procedure. No one, ever, will show that one oil is slightly "better" than another through a spectrographic analysis. When we ran UOA in college it was to infer machinery operating condition, never as a discriminator between which oil was better than another. It does not work that way. Sure, if you put water in the system and run it as "oil" then the analysis will show that. But not for two or more appropriately rated oils.You still did not answer the question. You stated "The iron difference cannot be attributed to the oil." And sorry If I came off Rude but why did that drop with the Amsoil same mileage same filter
Far too many variables in everyday driving for another thing. You test comparative oil quality through expensive, complicated and statistically valid methods. It's how you get results that matter.