@brokencase:
I knew I would draw a rebuke for citing "expert opinion" without solid data of my own, so no need to be sorry.
We must be careful not to confuse theory with fact, and fact with faulty data.
To make the prediction that aluminum "will" wear faster than stainless steel, you need a sound theory that explains all the observed facts. If you say a steel knife blade will wear faster than carborundum then I will believe you because otherwise I wouldn't be able to sharpen my wife's cooking knives, instead I would just wear oval divots into the stone (which is what happens too, just not as fast as the knife sharpens.)

But if you say an aluminum chainring "will" wear faster against a chain than a steel one will, then you need to be sure that hardness really is decisive in predicting wear to the exclusion of other factors that the theory doesn't incorporate. In other words, does the "hardness" theory explain the observed facts of chainring wear well enough to predict that the Al-alloy rings I bought last week will wear out sooner than the steel ones I could have tracked down instead?

If your own past experience is, indeed, that steel rings "have lasted" longer than Al, chain maintenance and replacement being equal between the two, then my second-hand report that the opposite is true should be cast into doubt, the product of faulty memory or biased observations. The difficulty is that very few of us have run chainrings of both types in parallel over the same type of riding over many years. If we used steel for our early years and then switched to aluminum once we could afford better bikes (as riders my age did), how do we know that the perceived improvement in durability is not due to improvements in manufacturing processes unrelated to hardness of material, to our own riding style and maintenance practices, or even completely imaginary? You won't find many people who have used aluminum for a long time and then switched to steel. Testimonials from those folks would be really illuminating, but still at risk of imperfect recall without careful records. (And yes, the use of steel for granny gears suggests that hardness matters, especially in a small ring with little weight penalty when made of steel.)

Finally, recall that the surface of native aluminum undergoes oxidation almost instantly to the mineral aluminum oxide, which unlike iron oxide is tightly adherent and very hard. Perhaps this is what protects the soft underlying metal from abrasive wear, anodized or not, provided the chain is of consistent pitch between the rollers?

I don't doubt that we use aluminum chainrings purely because they are lighter. And we accept a small (when measured in the grand scheme of things) penalty in shorter life,...or we find an unexpected benefit in slightly longer life.

End of musings about "how we know what we know."/

Well put, conspiratemus. I think that describes well where we are. We don't really know, and any theory is plausible but none is sure, because it would take a heck of a lot of measuring to know, and the benefit would be marginal anyway.

I agree there are circumstances where experience appears to run contrary to known theory, and upon further examination, there turns out to be some previous unknown phenonema at work.

This case is not one of those situations. The phenomena at work here is well understood and there is a body of knowledge that extends well beyond the the musings of a cycling legend. This is also not "rocket science".
We don't make ball bearings out of aluminum and we don't make roller chain out of aluminum FOR A REASON.
I don't need to perform the experiment, because the theory is so iron clad in this case. Stainless trumps aluminum with regard to wear.

I do agree with electrik's earlier post. Because chainrings have more links engaged, the the wear is distributed and we can get by with aluminum chainrings (and save a little weight in the process). But I must confess that I am looking at a worn big ring on my Scott CR-1 after 3 years of use. It's not cheap to replace this ring gear.
I'd rather replace chain than ring gears.

The small chainring looks fine, not because of some mysterious effect, but because I am more often in the big ring! I wonder if Barry made a similiar observation and concluded wrongly! Who knows, maybe the steel rings he was using back in the 50's was inferior.

Rest assured that a stainless ring will out endure an aluminum ring hands down, I'd bet my life on it.

The original poster indicated that this is a single speed/fixed application. Part of the beauty of these bikes is that the simple chain setups are low maintinence. Putting a stainless ring on makes them even more so, and this is one reason why they make stainless chainrings availible for these bikes.

Is every grade of aluminium easier in every mechanical way to work than steel?

Pretty much.