They're real and tracked, but they're also not the main reason why airlines aren't on time, or any sort of reason for lower cruise speeds.
We've got percentages for delays attributed to the National Aviation System (including those for reasons other than ATC understaffing, like congestion management) here[1], it's less than half of those attributed to the carrier, with a slight trend fall. That doesn't mean ATC understaffing isn't a problem (patching gaps in shift patterns is bad for a whole bunch of safety related reasons, for a start), it means that the author is dead wrong that airlines won't do anything to jeopardise on time performance and government must be the only bottleneck.
[1]https://www.bts.gov/browse-statistical-products-and-data/inf...
The author did not claim that the entire or the main reason for longer flight times was ATC delays. He wouldn’t be dead wrong if he had made that claim since as you note, ATC delays are higher than they were in the 90s. But he didn’t make the claim that ATC was the full and only cause of delays.
His conclusion is that there are a multitude of causes, among them, ATC staffing delays.
The author explicitly dismissed the idea of "greedy airlines" as not passing his "sniff test" whilst the data shows their own actions are a more frequent culprit for delays than aviation services (be they ATC shortages or congestion, which combined only account for <20% of the delays; unlikely to cause of a fourfold increase in long delays by themselves never mind be a reason for slower cruising speeds).
I doubt an economist a little less enthusiastic about the ability of markets to give people exactly what they want than Max would have missed the obvious dynamic that when airlines are competing mainly on price in a thin-margin capital-intensive industry they absolutely can capture market share (from paying customers, not just points collectors) by accepting the risk of degraded service.
We've got percentages for delays attributed to the National Aviation System (including those for reasons other than ATC understaffing, like congestion management) here[1], it's less than half of those attributed to the carrier, with a slight trend fall. That doesn't mean ATC understaffing isn't a problem (patching gaps in shift patterns is bad for a whole bunch of safety related reasons, for a start), it means that the author is dead wrong that airlines won't do anything to jeopardise on time performance and government must be the only bottleneck. [1]https://www.bts.gov/browse-statistical-products-and-data/inf...