Anecdotes are not useless. They can be inspirational and they can help illuminate possibilities and methods that are otherwise opaque in large-sample statistics. Anecdotes that are consistent with our best scientific understanding are powerful ways of communicating knowledge and providing inspiration for change to those who don't have the ability or inclination to evaluate the studies. In this case, the father is doing several things that large studies tell us are beneficial to health. These unmentioned studies suggest that the father has probably improved his quality of life. Most people I know are far more motivated by such anecdotes than they are by credible statistical analysis.
I disagree with your list of uncontrollable variables. All of them are largely controllable. Much of the medication consumed today is either outright unnecessary or can be reduced with lifestyle changes. You can control the air you breath by controlling where you live and when/where you are most active. You can lower your risk of cancer and many other diseases by lifestyle and environmental changes. The fact that complete control and certainty are impossible is no reason for inaction.