If the point of this metric is to gauge how well US schools are teaching literacy, we should only include people who went through the US school system. Which means we should exclude immigrants, both literate and illiterate, from the reporting.
I'd like to include that pools of U.S. born children exposed to children of immigrants in their classroom who have higher English literacy and proficiency may very well benefit from an almost osmosis-like effect. This would seem to be not countable under the metric you're suggesting.