I think the Flynn Effect should be ignored, and IQ tests shouldn't use normed values. If the point is to match intervention to ability, then there isn't overestimating taking place. We are just correctly increasing the group size of those that need intervention.
Let's say studies determined that the ideal intervention in math for a 7 year old scoring a 130 on the WISC-III is to have 1st & 2nd grade compressed into one year and to skip review at the beginning of the school year. In 1992, that would be ~2% of the kids needing 1st & 2nd grade math compressed. 20 years later and 7% of the kids are scoring 130 on the WISC-III, I'd contend that the compressed math would be valid for the full 7%.
Renorming and waiting for tests to settle in before research is done means that research and longitudinal studies will continuously show results and interventions matched to the wrong population set. The compounded result is even more bored children, unless the overal baseline curriculum is drifting its mean in line with the overall IQ gains.
IQ should be a measure not a competition.
