Recognize that once you're in the GT range, the error bars on scores tend to get pretty broad, so some amounts of "fluff" are quite normal.
Violet, Comprehension is very commonly a low score for Asperger's kids -- it has nothing to do with listening comprehension or reading comprehension, but rather, it's about knowledge of basic social rules, like what the right thing to do is in certain circumstances or why people act in certain ways. A high Comprehension doesn't prove that a kid doesn't have AS, nor does a low score prove that they do, but it's something that often goes with it.
Similarities is also sometimes low for AS kids, but again, nothing is proof one way or the other.
The factor structure of the WISC is rather funky, and I suspect that you're running into it. They claim it's a verbal vs. nonverbal split, but what it's really more like a split between crystallized intelligence (stuff you know, primarily tested on the VC side) and fluid reasoning (ability to solve novel problems, primarily tested on the PR side). Some GT kids (who are commonly perceived as "high verbal") are seen as GT because they know a lot, and they tend to have VC > PR. Some GT kids don't know as much stuff, but are really good at figuring things out, and they tend to have VC < PR. But the PR factor is complicated by Block Design's heavy load on visual-spatial skills and speed. So it's a bit of a mess. Key thing is to understand that the name of a test or subtest may not mean what you think it means (or even what the test publisher claims it means).
