I'm very skeptical of the claim that cognitive function declines after 24 years of age (NB I am 32 so take what I say with a grain of salt). Here's the original paper.
Consider the data collection process:
Thompson, et. al. collected computer game action data from 3,305 players, aged 16-44. They recruited from "various social media websites where StarCraft players are known to frequent". Each player provided a single replay file, which was parsed by the researchers. The mean age was 21.7, the standard deviation was 4.2, no word on the median but the histograms does skew towards youth. The data is overwhelmingly male (only 29 out of 3,305 players were female) and 43% were from the US.
First, the authors aren't explicit on how they did the recruiting. Did they make posts to Twitter accounts, which is incredibly time-of-day dependent, when older people may be at work? Secondly, the participants are providing the game replay data, which again is a confounder. Perhaps many are providing their best performances, games at which they win. Maybe they were "off"? The point is, with so few older people in the study (nobody above 44), the confidence intervals must be huge.
Finally, what population is their sample trying to describe?
The sample is almost entirely male and very Western biased. Not to mention using computer game statistics to infer cognitive decline seems dubious. Generally speaking, video game reflexes is probably a function of age to begin with. It's also a function of a lot of other things that the study doesn't even try to control.
Makes a good headline, at least.