I couldn't tell you what DYS would do with this data, but, some thoughts:
Did the examiner write it up this way, with the clinical observations about testing of limits using the standardized version in a language he knew better than English? If all of the requisite subtests were administered, it should also be possible to calculate a Nonverbal Index, and perhaps submit that as supporting evidence for the verbally-laden FSIQ/GAI being artificially lowered by having had only three months of English immersion at the time of testing. (In addition to those included in GAI, you would need Picture Concepts, Picture Memory, and Bug Search.)
It's also not wholly true that filling in the verbal items with questions from a publisher-endorsed translated manual is illegitimate. That's actually the test design for the Bilingual Verbal Abilities Test (derivative of the WJ used for non-English speakers). Though the difference in GAI obtained in this case is negligible, and might be partially explained by the cueing involved in re-administering an item.
