I just tried your suggestion but this, too, resulted in more yellow tarnish rather than less. I think what's happening is that I am removing some kind of plating, thus revealing an underlying gold/yellow colored metal.
One possibility, which I am now tending to favor, is that there are two kinds of National HRS type knobs:
--The older ones, used in National's own radios and often with intricately customized numbering/lettering on the skirts. I have a number of those, and I think they are some kind of *coated* metal, probably not alumin(i)um.
--The newer ones, sold at retail for use in homebrew equipment, with non-coated skirts made of alumin(i)um.
Here's a photo montage where I've illustrated this possibility: http://tinyurl.com/czhtt9v
If this is the case, I would tend to favo(u)r the newer knobs and perhaps that is why they are so incredibly pricey nowadays! It's the older ones that I've been picking up for peanuts.... perhaps there is a process for re-plating the metal on the old knobs, which would be very niceesthetically speaking -- and presumably very cheap.
This is all in aid of getting a uniform look for the controls on my homebrew equipment, and it happens to coincide exactly with most of the knobs in the ARRL projects pictured in the publications from the mid-1960s. Nothing urgent, since I can easily use any old knobs for the time being....
Brand-new alumin(i)um skirts could probably be homebrewed with a metal stamper, but presumably it would cost a fortune to get the stamp custom-made. (Measuring with my callipers, the skirts are .05" thick, 1.44" diameter, and slightly domed, with a small lip around the outside.)
One issue with the National knobs, as you can see, is that the skirt is electrically connected to the shaft, making the use of insulated couplers mandatory on "live" shafts" for anyone who might otherwise be tempted to trust a knob's own insulation!!
73 de Martin, KB1WSY