So now you're demanding that we use a different period for different data sets. It gets worse!
The falsification criteria simply demands that a 15+ year period of no statistically significant warming exists *anywhere* in the record during a period of ever increasing CO2.
It's like if I bet that you're an alcoholic if you drink more than 5 glasses of wine in an hour, but you say I'm cherry picking the time you're in the bar, and claim you spend most of your days sleeping and not drinking :)
You've been drinking, you've been caught, by every possible measure, and you *still* deny it :)
Look, you want to claim that NOAA 2008 is bogus, and not representative of the true hypothesis of AGW, fine - but realize that this kind of ad hoc special pleading, and dismissing of failed predictions is *exactly* the kind of thing astrologists do.
At the very least, given the evidence in all the temperature sets demonstrated, you have to admit that your mythical, as of yet unstated necessary and sufficient falsifiable hypothesis statement for AGW, cannot exclude arbitrarily long periods of no statistically significant warming with ever increasing CO2. And prima facie, the fact that you're now asserting that AGW can be true *without* statistically significant warming, makes your premise incredibly unlikely :)