If you're 12 years old and have an IQ in the 150s

Then you're ahead of the curve. But IQs over 100 simply aren't that interesting, outside of the handful of savants (and even those are heavily overstated). The sociological purpose of studying IQ goes back to the idea of eye and ear exams and general physical fitness tests. The point is to find kids who are suffering some kind of physical impairment and alleviate it, not to pick a few mega-minds out of the crowd and put them on pedestals. Are you underperforming the crowd because of your diet? We need to get you more food. Are you underperforming because of an injury? We need to get you medical relief and rehabilitative therapy. Are you suffering from a congenital disease? We need to provide you with care to accommodate your needs. Can you not read the blackboard? We'll get you some glasses.

Are you batting 140 in a field of 100? Good for you. Now piss off, you're not important.

The difference between an average IQ 14 year old and a 140 IQ 14 year old is extremely stark

By the nature of statistical analysis, sure. Even then, the action you take in this case is to simply provide the kid with the next rung of activities, consistent with someone more fully developed. That's easy to the point of being trivial and not terribly interesting from a policy perspective.

The difference between an average IQ 40 year old and a 140 IQ 40 year old is much, much less significant.

At the "mean standard" of adulthood, you should be able to operate independently as a fully mature adult. The ability to operate a little faster and more nimbly doesn't change anything for you. Its the folks who fall under the normative threshold that need attention, because they're the ones struggling with the existing societal layout.

If everyone needs to climb a staircase to come and go from the city center, and the stairs are 10 steps high, the guy who can comfortably climb 12 steps isn't any more interest than the guy who can do 10. Its the guy that can only do 8 that needs support. And the degree of the disparity is less important than clearing the base threshold, because tripping over the hurdle is more of a general social problem than clearly it by a little or a lot.

There are real policy implications to the "everyone in this cohort operates at an 80 IQ". But the folks making these claims more often than not don't actually want to implement public policy friendly to this cohort. What they want is to segregate them and exploit their perceived weaknesses. In the same way that we target the elderly with phone scams and spam quack medical cures at folks with chronic conditions, the conversation around IQ inevitably departs from "how can we make a more equitable and functional society" into "how can I benefit from others' disadvantage".

Implicitly assigning IQ by race, gender, and other phenotypic characteristics tends to be about justifying the degradation of conditions of exploitation. It has nothing to do with their actual intelligence, save perhaps that the more predatory and gullible will aggressively target this group.

very very high IQ children almost never become brilliant, world-changing adult thinkers.

We like to apply the IQ metric after the fact. So everyone in the top income brackets and the highest positions within public and private bureaucracies get to receive some "High IQ" or equivalent merit badge. Nobody is actually out there sitting Jeff Bezos, Donald Trump, Magnus Carlsen, and Ken Jennings down to confirm whether or not they're in the top 1% of pattern recognition solvers. (Carlsen pretty explicitly refused to take one, leading to endless internet speculation) That's in no small part because scoring high on pattern recognition tests doesn't tend to track well with social aptitude, and so "High IQ" doesn't actually lead to rapid climbs in bureaucracies.