A number of economists have speculated that the mystery of missing wage growth is explained by the fact that slack is higher measured by prime-age EPOP than by the unemployment rate. In fact, changes in prime-age EPOP actually exacerbate the mystery.
The mystery is that 12-month wage growth has been flat for two years now (and actual/expected inflation risen so real wage growth is down) even as the unemployment rate has fallen. 
But prime-age EPOP has improved even MORE than the unemployment rate has over this period. If prime-age EPOP is your preferred measure of slack, then you would be predicting even MORE wage acceleration than if you think slack is best measured by the unemployment rate.
Remember, what matters is NOT if slack is positive (which you would infer from prime-age EPOP) or negative (which you would infer from the overall UR). Less slack means faster wage growth. Not negative slack means faster wage growth.
(An aside: the employer wage setting power / low unionization hypothesis cannot explain the puzzlingly comparatively HIGH rate of wage/price inflation in 2011-12 when UR was ~9%.)
A separate issue, I would be skeptical of empirical models that try to explain stationary phenomenon (inflation) with non-stationary variables (prime-age EPOP). The fact that prime-age EPOP was much higher in 2018 than in the 1960s tells you nothing about slack.
has a nice chart showing the close relationship between prime-age EPOP and ECI increases from 1994-2017. But the relationship looks absurd over a longer period. economy.com/dismal/analysi…
The point is that EPOP is measuring much more than just slack. That has mattered a lot over longer periods of time. To view it as a good measure of slack in recent years you have to believe these trends are all of a sudden absent.
The wage puzzle is diminished but only slightly when looking at ECI which has shown more growth than AHE or the Atlanta wage tracker. I am agnostic on which wage measure is best and tend to prefer GS and other trackers that combine them.
In summary, I don’t think alternative measures of slack can rescue the Phillips curve. Alternative measures of wages help a bit but not a lot. May just be the concept is too poorly measured to be useful, see Box 2-9 in the 2016 ERP and @rortybomb today vox.com/the-big-idea/2…
