Another angle to this beyond the data being useless because of how it was collected.

High performance workers likely recognize that skills get them paid, so the opportunity to get better skills should always come before money, if you want to earn the most money.

Example: A young engineer is offered a job at MSFT at $100k per year being a cog on a team doing maintenance on an existing product. Or they can go to a startup for $60k a year doing cutting edge development of an innovative new product. Clearly the second opportunity likely has the best career benefits, and probably the highest career earnings.