What you have to remember when using the argument that young kids may be uncooperative/distractible/tired etc. is that all these conditions applied to the norming sample, too. If you take a sample now of kids testing, their scores will be "inflated" if they are on average less uncooperative etc. than the norming sample, "deflated" if they are more so. There's no reason that I can see to expect children being tested now to be *more* subject to those factors that might reduce their scores than the norming sample were, and therefore no reason to expect their scores to be underestimates as a group.
If we dispose of that set of arguments on those grounds, then what remains is the regression to the mean argument.
What an individual tester sees is going, of course, to be heavily dependent on which children come to that tester to be tested. That could be affected by a lot of different things. For example, a tester who draws testees from a population where high IQs are the norm and only a really exceptional child is likely to be brought for testing (or one, like Miraca Gross, famous for being interested in PG children) is more likely to experience that scores tend to be underestimates than a tester who tests from a more average population.
