I'm sure that others will weigh in and opinions may vary, but I'd say that, based on the scores you have right now, unless you also have achievement scores that more than meet the DYS criterion, he may not have the best chance of being accepted. That doesn't mean that he's not a good candidate for DYS.
We, too, had one less than stellar testing experience with one of my dds where the report had some truly offensive parts such as a written comparison to her older sister and opinions that flat out were not supported even by the testing the psych had just done (yes, she's above the 99th percentile, but she's not that atypical and you shouldn't push to have her in a GT program that is designed for the 95th and up, for instance).
When the rapport isn't there or the tester is less experienced with giftedness than s/he put forth, the numbers may not be the best estimate. That said, I think that you're going to have to have the numbers to show Davidson that he's PG and these ones probably wouldn't do it.
The tester should have calculated a GAI with the discrepancy btwn VCI/PRI and WMI/PSI, though. Typically that would be done using the standard tests for VCI & PRI and would give you a 136 (99th percentile). It seems a lot more likely that he is a 99th+ percentile kid than a 95th percentile kid (125) based on how you describe his academic performance.
However, you could theoretically get him awfully close to DYS criteria using some of the alternate tests. If you picked the highest numbers from the VCI & PRI tests that were administered, you wind up at 143, which is the minimum number to get into the 99.8th percentile.
The one thing that does stand out that made me think this may be why the dr was suggesting ADD is the fact that both the WMI and PSI were lower. That's exactly what my one 2e child, who is HG+ with ADD, looked like on the WISC-IV. None the less, if I didn't see the symptoms in the child, I'd be very hesitant to say the least to say that this is the reason. I'd consider it inappropriate to diagnose a disorder solely based on IQ spread unless there are other indicators. The IQ spread can raise questions, but there could certainly be other reasons for it.
