So to clarify, I actually meant that I -have- seen more extreme 2e profiles, but his is definitely up there.
I would definitely look into speech and language evaluation for some kind of expressive language impairment. Your observation that he struggles even with organizing oral language is a strong flag for true language impairments, not "merely" dysgraphia, and, I reiterate, probably a key contributor to his seemingly intractible anxiety. Telling a story about your feelings requires almost all the same skills as a generating a narrative about some other reasonably familiar topic.
The achievement testing narrative doesn't appear to break out the two components of scoring sentence composition, but your description of his oral language backs up the hypothesis that it's not just mechanics that was weak.
IQ tests should always be administered under standard conditions first, to obtain normative scores, but I generally test limits afterward, when it appears clinically indicated, as, although it is true that many people see a slight increase in scores with extra time, it is not actually true that most people would see a large jump with extra time. The reason the official time limits were chosen was precisely because, in the standardization sample, very few people were able to complete significantly more items when given additional time. Most examinees give an answer, and that's it. It's not like they keep looking at the items, figure out they were wrong, and change them to the correct answers, if you give them unlimited time. I've seen many, many students keep working for lengthy additional times, only to come up with the incorrect answer. That's usually what happens with additional time. Or they work for a while and give up. This is why I let students work through the time limits only when they either appear to be making legitimate progress, or when it seems to be in the best interest of maintaining rapport to let them keep at it (i.e., the student asks to try a little longer). Even then, I will stop them when it gets excessive. The outliers are the ones who keep reasoning their way through hard items and manage to come up with the correct solution after an extended time. The time limits are to expedite testing, and to reduce frustration for examinees who are approaching their natural ceiling.
So additional time might get most people a little further on the test, but not -much- further. As it happens, there is also a standardized way of testing limits, using the WISC-V Integrated, which has alternate response modalities and test conditions for systematically testing limits. Not every evaluator has access to this instrument, though. For instance, given what you've reported about his vocabulary acquisition skills, I'd be curious how he does on picture and multiple choice vocabulary measures (which are expressive language-reduced), vs open-response definitions. Or Block Design, which has an alternate clinical score (on the standard WISC-V) without bonus points for speed (still has the same max time limits, though), and also a motor-free clinical version (on the WISC-V Integrated).
On the counseling front, operating on the assumption that there is an expressive language impairment still unaddressed, it might be helpful to consider adding some therapeutic approaches that are not reliant on using complex language, such as art therapy.
