I have an appointment in ten minutes, so I can't comment at the length I usually do
...but was there any indication from the examiner (or could you obtain one) as to where they would estimate his FW score to lie if his overtime responses could have been included? How many uncredited correct responses did he have? Just one or two doesn't make a huge difference, but if we're talking four or five or more, that would be quite enough to account for the drop since last time (taking into account natural testing variation).
And secondly, the Index scores are much more reliable from test to retest than the subtest scores are. I wouldn't make any high stakes decisions based on a change in subtest scores alone, especially when the conditions also changed from medicated to unmedicated.
