Hm. Sounds like the nominal score has a decent likelihood of being a low estimate (with that many additional correct--but uncredited--items), but it also appears that the task was genuinely challenging for him. Your IRL description does align with weaknesses in cognitive flexibility, which additionally supports the interpretation of this as generally a real result (if somewhat lowered by frustration).
The directions for this test explicitly note that there will often be more than one way to solve the problem. It is also untimed, so should not be affected by his visual-motor speed, but it does use visual materials, so it's possible that that was a contributing factor (but then, FW and VP are also visual, and he did nicely on both). I think it's more likely that real underlying vulnerabilities in cognitive flexibility and abstract problem solving affected the MR result. That can have quite a bit of impact on his school experience. In your smaller, more individualized--and much cozier--homeschooling environment, I would expect that you have better facilities for adapting to his learning and emotional needs day-to-day, rather than forcing him to adapt to the rest of the class. This likely is reducing his overall stress level by quite a bit, which allows him to use all of his cognitive and self-regulation skills more effectively.
BTW, you might want to edit out his first name from your last post.
