Welcome, amanda!
(And thank you, indigo, for giving me a nudge!)
indigo has already posted useful contextual information about cognitive assessment.
I will add that understanding the concept of regression to the mean can be helpful. By this, I don't mean that some of the high scores are somehow false, but simply that it is extraordinarily rare for the same rare event to occur multiple times. Very high scores are rare events, affected by factors in addition to true intelligence (whatever that is). Multiple factors must align in order for low frequency scores to occur. Many of these factors become more predictable as children mature, and as their learning experiences become more uniform (e.g., through schooling).
Regression to the mean is also part of understanding how test performance varies on different tests (in addition to the actual differences in the test structures). When comparing tests (both cognitive to cognitive and cognitive to academic achievement), tests typically have correlation coefficients, which are one way of expressing how similar the results of the two tests (or readministration of the same test, in the case of reliability coefficiencts) are likely to be. An example: a high correlation between two gold-standard instruments (such as the DAS-II and WISC-V) might be in the neighborhood of 0.8. As a back-of-the-envelope approximation, this would suggest that a 124 on the first test could quite reasonably show up as a 119 on the second test--which is almost exactly what you see on the two verbal measures, which are the most similar across the two comprehensive measures (which isn't actually saying much, as both tasks on the DAS-II are different from both tasks on the WISC-V).
Leading to the more important consideration, which is that the areas labeled with similar constructs are actually sampled from very different tasks, in most cases:
The DAS-II verbal tasks have visual supports or manipulatives, and include a picture vocabulary task and a task of following oral directions (which is extremely sensitive to fluctuating attention). The WISC-V verbal tasks are both purely verbal (for the most part).
The DAS-II Nonverbal Reasoning has one task that is much like one in WISC-V Fluid Reasoning, but the other tasks are quite different from each other, with the DAS-II task more of a visual analogies task, and the WISC-V more of a quantitative reasoning task.
The DAS-II Spatial tasks do share a similar task-type with the WISC-V Visual Spatial, but then there is also a relatively complex fine-motor/pencil skill task, versus and motor-reduced spatial thinking task on the WISC-V. So the DAS-II has a fine-motor involved task and an even more fine-motor complex task, and the WISC-V has a fine-motor involved and a motor-free task. You can see how someone fine motor speed is assessed a bit lower than there highest area of strength might be at a bit of a disadvantage on the DAS-II. (Note the WISC-V PSI is actually about the same as the DAS-II SC.)
And then, of course, the WISC-V has the WMI and PSI, which have no analogs when computing the DAS-II GCA.
Plus, the NNAT-3 is essentially like one DAS-II Nonverbal Reasoning or one WISC-V Fluid Reasoning subtest. And splits the difference between those two not-entirely-the-same index scores.
And then to throw in an entirely different twist...if your DC is in fact dyslexic (which your neuropsych says she is), then you may see even more fluctuation over the years as her reading vocabulary alternately lags or catches up to her "true" verbal cognition. In order to minimize these types of impacts, and to maintain appropriate cognitive challenge, I would highly recommend that, in parallel to good phonetic-based Orton-Gillingham-based reading intervention such as Orton-Gillingham itself, Wilson, Barton, or one of the open-and-go home-based programs (e.g., All About Reading, Logic of English, Nessy.com), you continue to expose her to high-level vocabulary and language through any oral means available, such as conversation (obviously!), high-interest videos on topics of her choosing, and audiobooks at her listening comprehension level. If she enjoys writing, let her use whatever modality of expressive communication poses the fewest barriers to her language formulation (dictation/scribing, speech-to-text, typing, etc.) especially to do self-selected writing, so that she has maximum access to using and developing language and vocabulary at her native level.
