alt Hacker NewsI don't think basic color accuracy matters for this, it's more macro. In other words, two professionally taken images of a painting aren't going to make it look bright and colorful in one, and dark and somber in another.
Whether there's a slight green tint, or a certain blue doesn't pop quite as much, doesn't seem like it would affect the findings.
I've seen exactly what you said isn't going to happen in two pro photos of an old (pre-1800s) painting.
One might have been altered to reduce the effect of centuries of oil-lamp soot, but it's still true.