Earlier quoted context omitted.
That's the whole point: it doesn't matter why they died. If you're doing things well, excess deaths are low. If you're doing things badly, excess deaths are high. Could be a multiplicity of reasons that connect "bad" and "excess deaths", and the specifics won't matter if you're looking for a summary of "normal human life, good living standards, capacity to adapt".
Of course it matters why people die. If I was an analyst in the early 1940s analyzing why there was an excess of human deaths and wrongly attributed it to an increase in bear attacks, wouldn't that be really dumb? If you're trying to solve a specific problem, isn't getting the cause of death wrong a really big deal?
Misattributing it to bears when it was really lead in gasoline or plastics in the air or whatever really caused it just delays your ability to help fix it, but that's a different question.