There's a statistical issue with the study he quotes, I didn't dig deep enough to find a conclusion:

Is the study longitudinal, ie does it follow gamers as they get older?

You want to do that, because you don't want to confuse the issue of what cohort the gamer belongs to. You want to differentiate between "people like games less as they age" and "people who were born in 1980 like games less than people born in 1990".

In my own experience, once you have kids, that's it. You can't invest in any game that takes time to learn. Before kids, you could have a 2nd job healing or tanking for 3-8 hours every evening after your real life job. Or working on your shooting skills. Or watching RTS videos.

Now I can do casual gaming, which works when you're commuting, but it's not nearly the same commitment.