Hm. What you are describing sounds more and more like executive function weaknesses, which also fits the questioning about ADHD, in addition to the existing EF results on the ENFEN and CAS. Aspects of EF that these behaviors would align with:
1. planning & organization: odd mistakes on layout when copying text, as well as difficulty inferring directions.
2. cognitive shift: adjusting slowly to changes in format or task demands.
3. sequencing: not knowing where to start from or what she should do first. This could also be from weaknesses in another EF skill (initiation).
4. inhibition (which is the positive skill that is impaired when we talk about impulsivity): inconsistencies in performance, especially on multiple choice tasks.
5. probably a combination of EF skills, but likely resulting in impediments in cause and effect thinking, which often shows up in inferential reading comprehension.
Some of these would have impacts on visual working memory, which (of the kind we've been discussing) is sequential, but not on simultaneous visual spatial memory.
Along those last lines, I also wonder whether she is a simultaneous or gestalt thinker, rather than a sequential thinker. That might explain why she did so much better on visual spatial tasks when young (when the designs are much smaller, and don't require as much part analysis), and was only average a few years later (when you need some part-to-whole thinking to efficiently reproduce the designs). Note the Rey-O results describe her as starting from the big picture and trying to work her way down, but missing some detailed elements.
And yes, the behavior with card-flipping memory games and with scripts (as well as the Rey results) suggests that she is extremely efficient at transferring to long- or mid-term memory, which isn't necessarily reflective of any working memory issues that might exist. Retrieval, on the other hand, may not be as efficient for every method of accessing from long-term storage, which may result in inconsistencies in her ability to demonstrate learned knowledge and skills, such as those you report.
I wish we could do some more fine-grained item analysis, and see if her Matrix Reasoning performance was slanted toward errors in the sequential items or the gridded items. Or see how she does on some of the other cognitive tests, like the KABC-II or CAS, which explicitly assess simultaneous vs sequential thinking. I'd also be interested in more thorough assessment of her memory structure, such as through an instrument like the WRAML or CMS.
As far as skill-building: well, EF happens to be one of the cognitive skills that can be taught explicitly (perhaps because it's actually a relatively late-developing skill, so the window for brain plasticity is longer). Not so much through an overpriced "braintraining" computer program, but through everyday parenting and classroom strategies, modeled, explained, scaffolded, repeated, and reinforced over time. For example, she might do well with mind maps or graphic organizers to scaffold the organizational process for longer writing assignments. Or scripts and flow charts for specific problem solving algorithms or procedures. Or visual models for the first few times she does a new task-type (or format for written assignments, etc.). Or explicit instruction for strategies for getting unstuck, such as "backing up", asking oneself framing questions (e.g., "what am I trying to find out?" "what do I already know?"), turning the problem around (various ways of looking at a question from a different angle, sometimes literally, such as with math problems that have diagrams).
If you haven't seen it before, I'd highly recommend Peg Dawson's "Smart but Scattered." (US Amazon page:
https://www.amazon.com/Smart-but-Scattered-Revolutionary-Executive/dp/1593854455/)
FWIW, my lowest-EF kiddo is now (after years of parental scaffolding) a very organized young adult, with sufficient self-regulation skills. Still not exceptionally strong sustained attention or working memory, but compensated by long-term memory and high conceptual thinking. DC has also learned that having to organize and understand information enough to teach it to someone else is what transforms new learning into permanence. Which also makes them a popular classmate!
