[Someone]

I don’t see your first link supporting the claim “physical activity has actually risen significantly in recent decades”. When I click through that first link, I eventually get at https://journals.humankinetics.com/downloadpdf/journals/jpah..., which is about self-reported physical activity and thus “subject to recall and social desirability biases.”

It also only is about leisure time physical activity. So, if a subject buys a robot lawn mower, stops mowing the garden every week, and starts driving to the gym to do half an hour of moderate exercise once a month, the number measured here goes up.

Ignoring those, it says

“The prevalence of insufficient activity was not significantly different in 2018 compared with 1998 for most subgroups (Table 3), with exceptions of increases among men, adults aged 65 years or older, adults of Hispanic origin, adults with less than a high school education, adults in the Midwest or South Census regions, and adults with obesity, and a decrease among adults with a college degree or higher.”

and

“recent increases in meeting or exceeding the guideline overall are primarily driven by more people reporting sufficient activity to meet the high guideline, not the minimal guideline.”

So, it seems any increase in exercise comes from those already doing it doing more of it.