It seems the criticism is indeed Berkson's Paradox, but the example is different to the canonical example of Berkson's paradox.

In the canonical example, you have uncorrelated attributes, eg skill and attractiveness in actors, forming a round scatter plot with no correlation. Selecting a subpopulation of top actors who are either skilled or attractive, you get a negative correlation. You can visualize this as chopping the top-right of the round scatter plot off: the chopped off piece is oriented in roughly a line of negative correlation.

In this example, if you look in the linked paper inside the post by Dimakis, there is a positively correlated scatter plot: You can tell the shape is correlated positively between youth and adult performance. But in this case, if you condition on the extremes of performance, you end up selecting a cloud of points that has flat to slight negative correlation.

Here's the wikipedia on that one

https://en.wikipedia.org/wiki/Berkson%27s_paradox

Berkson's Paradox seems to rely on the selection criteria being a combination of the two traits in question -- in the example I keep reading about, only "famous" actors are selected, and actors can be famous if they are either highly talented or highly attractive. But in TFA, surely the "high performance" selection filter applies only to the adult performance level?

To put it another way: If selection was restricted to people who performed highly in either their youth or in adulthood (or both), Berkson's Paradox explains the result. If selection was restricted to people who performed highly in their youth, or if selection was restricted to people who performed highly in adulthood, Berkson's doesn't explain it.