Colorblindness in heterozygous females?

Nelumbo, of Biology Educators, has a good genetics question:

1. Red/Green colorblindness can be a x-linked trait.
2. If my husband has this type of color-blindness, our female child would have his 'mutant' X chromosome.
3. Let's assume I do not carry the trait and donate an 'normal' X chromosome for the trait.
4. Our daughter would be heterozygous for the trait, or a carrier.
5. One of each of her X chromosomes would be randomly inactivated in each cell

so....

Would about half of her optical cells show the normal protein? Or is the inactivation early in development, so all her optical cells would be the same ( either all maternal or paternal X inactivated)? Or even weirder yet, could she be colorblind in just one eye?

I'm no expert in developmental genetics (PZ, where are you?), but I can at least take a stab here. X-chromosome inactivation is the name of the phenomenon wherein one copy of the two X chromosomes found in a female's cells is inactivated by turning the chromatin in the chromosome into heterochromatin, preventing gene expression. X-chromosome inactivation occurs in both insects (e.g., Drosophila) and mammals.

To answer Nelumbo's question, the matter of when the X chromosome inactivates is important. If the X chromosome is inactivated early in development, then large patches of cells should have the same X chromosome expressed; if the X chromosome is inactived late in development, then each cell might express a different copy of the chromosome than its neighbor. I dug out my old developmental biology textbook (Gilbert 1994), brushed off the dead Drosophila, and found this experimental answer to the question:

One of the earliest analyses of X chromosome inactivation was performed by Mary Lyon (1961), who observed coat color patterns in mice. If a mouse is heterozygous for an autosomal gene controlling hair pigmentation, then the mouse resembles one of the two parents or has a color intermediate between the two. In either case, the mouse is a single color. But if a female mouse is heterozygous for a pigmentation gene on the X chromosome, a different result is seen: patches of one parental color alternate with patches of the other parental color.

So, the X chromosome inactivates early (within 10 days of the start of gestation in mice), and thus large patches of cells have the same X chromosome inactivated. The same type of system is responsible for the pattern of female calico cats.

This means that it's likely that large patches of Nelumbo's heterozygous daughter would have the same X chromosome expressed, and thus it seems at least possible that her daughter could be at least partially colorblind. This conclusion is backed up by this website:

Red-green colour blindness is hereditary and is passed via the X chromosome. Women, who have two X chromosomes, are usually not colour blind. But about 15% of them are carriers (i.e., they inherit an X-chromosome carrying an abnormal photopigment gene array from one parent) and may share in part in the colour blindness that they pass on to their sons, owing to a process of dosage compensation known as X-chromosome inactivation of lyonization.

This wikipedia entry also confirms that X-chromosome inactivation does change the color vision of carrier females; it implies that there is variation in the receptors present in each eye, and thus the female shouldn't be completely colorblind. However, I haven't found a good reference clearly explaining the range and frequency of clinical symptoms observed in females heterozygous for red-green colorblindness.

So, I leave it up to the rest of you great geneticists (and medical folks) to provide more details.

Reference:

Gilbert, Scott F. 1994. Developmental Biology 4th edition. Sinauer Associates, MA.