You don't need to be Wilhelm to conduct your own testing on a substrate, put it out in the sun either in a window or just directly exposed outdoors for 8hrs a day. Use a spectro to record changes after each session. Changes will occur within a few days to a week, NOT years.
UV top coating does little to "protect" against this change. It's like you wearing sunscreen at the beach for 8hrs... You are still getting exposure and are likely to end up burnt. I have tried many of these UV coatings and the result is the same. A piece with a top coat is spared little more than those without.
But hey, don't take my word for it... Just do it youself.
It seems the only way this "sun" test would provide any useful information is to do it simultaneously with a product that has a "known" permanence. Using the sun is a form of an accelerated aging test, just not as controlled and measurable as standardized tests. Of course these tests are still somewhat theoretical ... it could be that "accelerated" aging tests also somewhat accelerate aging anyway, and images not subjected to such extremes will actually age much slower. If you use the sun and it fades in 2 months, is that good or bad? Is the test at the equator or in Alaska ... considerably different circumstances. Could be 2 months is the equivalent of a 100 year rating under controlled tests such as Wilhelms. Perhaps if you test a paper and ink combination that has some supported research at the same time, you can tell if it fades faster or slower ... at least you have a base for making a comparison.
One thing I've always felt is the pigments in the ink are what actually fade. While the medium itself can certainly contribute to fading more rapidly, any well made medium should still perform very similarly. A well made canvas which perhaps hasn't been tested may actually have the same ink receptor coat as a well made paper that has. Unless there is a problem with the canvas itself the canvas should perform very similarly to the paper, and in general the ink set itself should perform similarly on most media that is at least somewhat archival in properties.
As far as the ability of a coating to absorb UV and slow fading, I believe there is some evidence that it does make a difference. part of the reason however isn't due to absorbing UV, but also protecting the pigments from various gases which over time can be damaging as well. Currently there is some dispute over which gases should be tested, and I've read that some even believe Wilhelm needs to consider measuring some additional gases and their affect. There is some discussion that for long term fading, gases may be the more damaging component.
My comment on the UV absorbing coat was more about the UV not being able to excite the OBA's in print. If you absorb all of the UV light before it reaches the OBA's, would a reaction occur, or would enough UV light fail to hit the image to excite the OBA's, so the modification of UV light to visible light doesn't happen. I've never tested this with coatings or with glass for that matter, just read about it. Some comment that if you frame a print under high quality glass the OBA's aren't apparent, so the print won't actually "yellow" over time since you can't see the effect of the OBA's. Just some things I've read and seem logical ... guess I should actually try it sometime and see for myself.