If every single person in the country made exactly the same salary, and saved the same percentage of their income to cover the exact same modest retirement life

If every single person in the country made exactly the same salary, and saved the same percentage of their income to cover the exact same modest retirement lifestyle, you'd still have people (close to retirement) who appeared "rich" and people early in their career who appeared relatively "poor". Using wealth data without normalizing for age you could also make it look like there was a vastly unequal society, and of course there would no policy you could adopt to "fix" the problem that wasn't there. What the author is doing here is literally the definition of bad data science.