I think this line of thinking is just sloppy. "Theoretically" the efficient market only gets to pareto optimality, i.e. we can't make the commuting low-wager worker's situation better without someone else being at least slightly worse off, but there's no lower bound on how bad things can be for that low-wage worker. Nothing assures us that we converge to a state where at the low end "the wages should come up to make it possible to live [in SF]" or anywhere particularly close.

I find your thinking sloppy. There, was that helpful?

"better without someone else being at least slightly worse off"

Exactly. The point (if you go back through my comments) is that there seems to be a demand and supply imbalance based on the skill/type of labor. We should be looking at ways to slow higher end wage growth and promote it in the lower end. One way to do this could be around education. Essentially, increase the number of developers on the market while decreasing the number of people who are only qualified for the lower wage work. We've kind of seen this with the pandemic shrinking the labor force and driving those lower end wages higher.