Again, I'm not saying that you can't see a visual difference. It's a matter of being able to
accurately and consistently assign a label of "stainless" or "carbon" to different samples of steel by sight when in a state of bright finish. If you took 100x Opinel blades, with half being stainless and half being carbon, mixed them all together in a tumbler, and then had even a very experienced person sort them into two piles, with the piles scored on both the metrics of all the blades in a pile being
the same kind and on correctly identifying
which kind is stainless and which is carbon I think that people with a keen eye could do very well on the former task (sorting one type from the other) but only score statistically equal to the 50/50 chance they'd have randomly guessing which one was stainless and which one was carbon. Blindfold the person and it would be totally random in both tasks.
In laminated scythe blades I can very commonly identify the separate layers of steel and iron, and you can see a photograph of it at 7:24 in this video showing a laminated blade that was deliberately ground too much on one side, making the edge cladding iron. The ribbon of paler coloration in the blade is the high-carbon cutlery steel edge layer, while the cladding is high grade Swedish iron.
But if I didn't tell you what material was what, and the faint flecks of storage rust on the factory bevel were removed, you wouldn't be able to tell on appearance alone that none of those three layers are stainless. You'd only have the context of it being a vintage scythe blade etc. to give you that info. A cropped image with everything else taken out and other context removed, you'd just be able to tell it was laminated, and not really anything beyond that.
So, to reiterate, you can see that they are different from one another, but you cannot use that visual difference to reliably improve your ability to guess correctly above the average for completely random guessing. Two different carbon steels can look different from one another. Two different stainless steels can look different from one another. Some carbon steels and stainless steels could plausibly even have the same overall look to them in luster and tint. If you were to take a random spread of steels, carbon and stainless, and prepare their surfaces in the same fashion, and have someone match the names of the steels to the appropriate blocks, the results would possibly even be worse than random guessing. Does that clarify my position at all?