I thought IQ was specifically designed to be normally distributed?

That’s true.

Also, this is where theory meets reality.

Defining something as “the average” or “one std deviation “ is strange when the population is unmeasured and changing.

There is no assumption, that is the definition of IQ.

Is human intelligence a normal distribution? Probably not. But IQ is extrapolated as such, and probably useless anyway. Which make absurd claims like this even more laughable.

IQ is offset and scaled with the goal of producing that normal distribution, but that process is informed by the available data. Outliers don't get forced onto the curve, because there definitionally isn't enough data to figure out how outlying they are as a percentile.

Exactly.