XRF hand held instruments are not as sensitively accurate as many may think.
I note that the acceptability range for the particular instrument used in this case is in the range of 89.8% to 90.21% for the gold readout.
That is OK, to my way of thinking.
Ru could well show up in the analysis, but hand held instruments may well detect any element but the accuracy for some hand helds could well be in the range of about 0.2% to 0.4%, and still be acceptable for that particular instrument.
The other factor is to what degree of purity the refiners (or the MInt) wish to refine their gold, and still be acceptable to meet the required purity standards suitable for the gold coinage.
Other than make a general comment in this regard, I do not know enough in an accurate and absolute sense about US standards for acceptable purity of the gold.
------------------------------------------------------------------------------------------------------------------------------------------------
For really accurate atomic (and much more expensive) analysis, mass spectometry is needed.