I think there’s ample evidence that seatbelts and airbags save lives, thereby improving lifespan as compared to the control (same number of car trips).

Similarly, avoiding or ceasing of habitual smoking of cigarettes has ample evidence for lifespan improvement.

Can you speak more to "Also there can't be scientific evidence that anything improves lifespan." I am missing the subtlety, but I would think statistically with twin studies and controlled experiments you could?

Of course there can be evidence.