> When companies like Google see their healthcare costs drop by $12,000 per employee annually and productivity increase by 25%, we observe a restructuring of corporate America that makes remote work a minor adjustment.

In my entire office, there's one overweight person that I know. Out of thousands of employees. There could be more, and other cities might be different, but you get the idea.

The author intentionally or mistakenly applies some numbers to whole populations.

Keep chaining impacts while eliminating small coefficients, your end result will be off by more than a few orders of magnitude.

EDIT: I need to be more clear about my point: My BigTech office does not represent my country stats: Canada has a 26% obesity and 36% overweight adult population. Just taking those numbers, applying them to all populations is wrong.

The analysis is wrong in other subtle but important ways as well. A 25% increase in performance of an allegedly low performing group of a company would not increase the overall company's efficency as much.