Welcome!
In general, one should start from the FSIQ, which has the strongest psychometric properties. From that standpoint, these data are consistent with GT identification.
With regard to the profile, it may be helpful to look at the range of index scores obtained, which is actually not as hugely disparate as might first appear. From top to bottom, the indices span 16 points, which is just barely over one standard deviation. And the clustering is not really between GAI and CPI, but is a relative strength in WMI. Look at the other three indices, which span only 6 points among them.
Now consider the subtests and how they appear to cluster. Within the PRI, there are two average subtests (VS) and two Very Superior subtests (FR), as you note. In fact, if you break out the skills, what you see is visual spatial skills in the upper end of the Average range (this is the most like a relative weakness), verbal comprehension and processing speed in the Superior range, and fluid reasoning and working memory in the Very Superior range. PS is not unusually high compared to the others; it's actually nearly identical to VC. And WM is not unusually high; it's nearly identical to FR.
These latter two areas (FR and WM) happen to be the two cognitive clusters most closely associated with math achievement. (In fact, the WISC-V combines FW from the FRI and the analog of Arithmetic from the WAIS-IV WMI into a secondary cluster called Quantitative Reasoning.) Sometimes the VSI has relationship to math achievement, but with such strong abstract thinking skills, there are likely to be multiple cognitive workarounds for any relative limitations in spatial reasoning skills. I would say this profile predicts a learner who is decently strong in language-based thinking skills, but even better at math, especially abstract math.
In short, this is reasonably interpretable as a GT profile that is not particularly unusual for a learner who is gifted in math.
