As someone who's gone through hired's process last year with a close friend, I feel this does a disservice. Primarily because COL is never 1:1 in spite of the very best analysis available and I feel like this becomes a credible source those in hiring positions. The train of thought goes something like this, seemingly from anecdotal evidence from my close friend: "Oh you're making $70K in Atlanta as a PHP developer? That's like $150K in SF!" (the SF value was plucked out of thin air because I don't quite know the calculation). When the reality is in Atlanta that's well below market rate for a seasoned developer. This is also the state of their data compiled from last year but I suspect more than most will treat it as a projection or try to run the increase/decrease as one.

My understanding may be seriously flawed because I view salary negotiations as nowhere close to a zero sum game it should at least attempt to aspire to be. From the get go, everything is entirely skewed in the employer's favor and this just feels like more ammunition.

My close friend and I shared the train of thought that if a company based in SF has $x for developers, while they can likely get more remote developers for the total at their COL decrease, splitting some of the difference with those employees would go quite a long way. Employees likely wouldn't see that total compensation range anywhere else in their current location and that difference becomes its own motivator, at least for someone like me who has struggled to reach market rate. It becomes a double edged sword when that offer is vastly above market rate as finding another job in that range would be increasingly difficult. Employees would likely be way more tempted to stay well after burn out sets in because what should be closer to the norm becomes more of a golden ticket.