that seems weak .. so lets take two retirees, both with these same comorbidities etc, except one is still strong enough to walk around everyday while the other gets long winded and can hardly manage walking to the carport .. one would already expect them to have vastly different mortality risks, whether for covid or everything ..

basically it's a huge confounding factor in almost all these studies whether the comparison being made is simply between people who choose to exercise vs those who choose to be sedentary, or whether it is between people who still have the ability to be active vs those who are already weakened and homebound etc .. especially when you include elderly in such studies

(that said, ofc there are numerous better structured studies that show the clear benefit of exercise, certainly no doubts about that in general)