James Woods has a reported IQ of 180

[muriel_volestrangler]

79. Yeah, IQ is measuring "in the top x.y% of the population"

and at the extremes, there's no way to say a test has been able to sort people, even if you accept that what's measured is a permanent ability rather than how well they did on that test on that day.

For every 15 points above 100, it's one more standard deviation above the median, in a normal distribution. Most tables don't bother going as high as the 5.3 standard deviations that 180 represents, because the fraction involved is so small that it's meaningless in most real-world statistics. The fraction outside 5.3 sigma on both sides is 1 in ten million - so that's 1 in 20 million with a supposed IQ of 180 or above. So Woods is claiming he's one of the 16 most intelligent people in the USA.

160 is about 1 in 32,000. That is, as you say, about the believable limit of what tests might be able to measure. It would be the 10,000 most intelligent people in the USA.