I would agree that there needs to be a recognized measure/metric/barometer...to assess final results in order to get an acceptable P-level if that is what it is termed/needed...
How can Kaplan-Meier take into account the exceptional reasons for an individual passing...such as a predisposition left up to genetic makeup (family tree)...environmental causes at the early stage of life.

Job loss is predicated on location...city vs rural vs state...area population/ industry/field of endeavor...the degree of education after job loss...other factors might include outsourcing...competition with a better widget.

Time to failure is based on metallurgical makeup of part...known tensile strength etc...
I don't see these... as a true comparison to life and death.

Perhaps leaning more into (Mantel-Haenzel Test) and approaching it from a numbers only lens...imho