It looks like he had two clusters (6 subtests) done at age 9. The corresponding clusters on the WISC-V are the VCI and a combination of the VSI and the FRI. There are several factors that could affect his test scores without getting into whether his skills or capacity have changed in any way:
1. Norm obsolescence: The WISC-V has brand new norms. Your son was assessed (presumably) somewhere between five and seven years ago, when the WISC-IV norms were already about seven years old. (The upper limit for age on the WISC is 16-11.) This alone would be expected to explain a few points difference in the center of the bell curve, and possibly quite a few more points at the extremes, as the Flynn effect has been observed to be of greater magnitude at the tails.
2. Change in test structure: The WISC-V is composed of five clusters, rather than four, with the WISC-IV PRI roughly split out into the WISC-V VSI and FRI. Each of the new index scores is derived from one of the familiar WISC-IV-type subtests, and a new subtest (VP for the VSI, and FW for the FRI). One of the three WISC-IV PRI subtests has been removed (Picture Concepts). This makes the old PRI difficult to compare directly to anything on the WISC-V, although I usually expect some relationship to the VSI and FRI. (WMI is also notably different, but since he didn't have that previously, I won't get into that.) Each index score is now derived from only two subtests, vs three on the WISC-IV.
3. Differences in task expectations by age: Bonus points for speed are more important for older students than for younger students (such as on BD), especially for spreading the upper half of the curve.
Now if we are trying to make this a comparison with as close correspondence as possible, we can narrow it to four subtests: Vocabulary, Similarities, Block Design, Matrix Reasoning.
1. Two of them generate the comparable index score (VCI). This index score appears to have fallen a bit since age 9, but not so far that it cannot be explained simply by norm obsolescence and standard error of measurement. I would not be concerned about this one at all.
2. Block Design and Matrix Reasoning, if the old GAI is to be believed, have fallen significantly. The associated new VSI subtest (VP) is on par with the two verbal subtests. The new FRI subtest is at the same average level as MR. These two items together suggest to me that the VSI aspect of the old PRI has probably been lowered mainly by speed/bonus point factors. You might be able to test that hypothesis by asking for the BD-no-time-bonus process score, which is normed for accuracy, but not speed (it still has time limits).
3. Another hypothesis that might be worth exploring arises from the two FRI subtests. One of the issues that comes up at high school is the increase in abstraction, not only in the sense of concepts, but in the lack of obvious relevance. Your son reports that he doesn't do well because he "doesn't see the point". If this testing is an accurate representation of his cognitive profile, this statement may be an observation about his learning style, rather than about adolescent rebellion. It is possible for an individual to be extremely bright, but struggle with tasks that are too far removed from personal relevance. When skills or concepts are contextualized and personally-relevant, they may display exceptional performance; in the absence of that context, performance can fall significantly. I've evaluated students like this in the past. The subtest that was removed from the WISC-V FRI is the one with the most everyday meaning and natural context, and is the most amenable to verbal mediation (employing his strong verbal skills to compensate). Elementary-age students as a population are more concrete and context-bound than adolescents are, which makes it easier for a student who is not as strong in decontextualized skills to land higher in the norms.
And, of course, there may be other factors affecting his test performance, such as lack of engagement in the testing process, fatigue, emotional factors, etc.
