70% of the US working population are full time workers.

30% of the working population are thus excluded, and they are likely all (with some extraordinary exceptions) below the median wage. Which means that 60% of below median wage workers are being excluded (or 51 million people), which is a significant amount compared to the top 1000 people in the above example. Doesn't say so in the wikipedia article, but easy to tease out when they start weaseling with "fully employed" when a significant amount of the working population are not part time employed by choice.

It's only that high if you narrow the scope to only full-time and employed for the whole year. This neglects how much (or maybe rather how little) the us values a large proportion of its labor. I didn't check the math on the percentage they gave, but it's at least in the right ballpark from some very rough napkin math- 60% of below median is 30% of total, and if we make a not amazing assumption that the "curve" of earnings is linear, then we can just shift the median by the given 30%, which if you shift up from 47960 you get 68.5k, or if you shift down from 60070 you get 42k.