So if we limit it to arthritis it is interesting.
First, we have to understand that correlation is not causation, and anecdotal evidence is not quantitative proof. Just like someone who has cardiovascular disease suffering a heart attack and dying at a marathon is not proof that running is unhealthy, someone who runs having hip resurfacing is not proof that running is the cause.
To a point, you're right. There are absolutely trade offs, and the little evidence that is out there on this suggests that there may be a u-shaped curve. What we don't know is when you start to see that inverse relationship in terms of long term bone and joint health.
The tricky thing is that it actually may not be running causing the issues if there is in fact an inverse relationship. It's actually quite possible that previous medical history or genetic predisposition is the cause of the development of arthritis. Or it's the result of diet, lack of recovery, history of injury, etc.
There is a reason that current clinical advice is to not dissuade runners from running high volumes for fear of arthritis. There simply is not enough evidence to suggest that it poses a true risk for the majority of people.
Finally, there is some really interesting research examining both older non-professional high volume endurance athletes (runners, cyclists, triathletes) as well as former olympic endurance athletes. The findings suggest that these people who exercised at extremely high volumes and intensities for so long are actually healthier than most average athletes or sedentary people. There isn't an increased incidence of arthritis, CVD or mortality. Plus, more and more evidence from the ultra-endurance side of things is showing that if runners take time to recover after hard efforts, try to avoid stress fractures and similar injuries, there isn't a limit to how fit you can get or how much you can run without hurting general long-term health.