You generally find that when trying to color-correct a positive image with a heavy color bias is that you are missing a lot of color information so the end results tends to be a bit desaturated once you pull the color bias out. This is true whether it is an orange-looking image or a blue-looking image, the only difference is that if all you care about is fixing fleshtones, at least with the orange image you have a bit more red information to work with, which is missing in the overly blue image. Fixing an orange image in a print tends to lead to a bland skin tone but fixing a blue image in a print tends to lead to a dead grey-ish skin tone (which you can then warm up... but you end up back with a bland brown-tan color by adding orange over grey.)
But in terms of noise problems, it's a problem either direction and since with film, blue tends to be the grainiest color, it's better if the negative has excess blue exposure rather than excess red exposure in terms of making corrections. But again, this assumes working from the original negative, which has a lot of information to work with.
With a print, just like with Rec.709 video, what you can see in the image is all you have to work with, there isn't really any "hidden" layers of information.