It’s the human condition
It’s the human condition, safety culture, passenger reaction.
Fines for errant passengers can be countered by legal argument after the event.
During an event, there is a personal balance of perceived risk - humans are naturally risk takers. The value of the bags (at this moment) is more than a defensible fine.
“Danger of Death” or $1000 fine, psychologically illogical messages, but often used in ‘blame and train’, litigious rule based culture, a belief that human performance can be controlled.
An alternative situation, as promoted by many cabin crew, involves when to change behaviour - ‘when the fire is too hot, smoke too thick, water too deep’: its a personal judgement, in this case the passenger.
Safety assessments might consider this aspect. There may be little or no evidence of death directly due to cabin bags (historical risk), vs perceive likelihood of death from a known hazard, a subjective fear (projected risk).
The issue starts way back. Regulators allow more carry-on bags, commercial demand - what were the safety assessments based on at that time. Grandfather rights again; also note recommendation 22 in RAeS report - bags.
If evidence from recent events warrants reassessment of the safety case, then solutions might consider restricting number of passengers on board, revised requirements for timed evacuations with pax and bags, changing the aircraft design; all with commercial implications.
So the question is not ‘what is the risk’, but who holds that risk; regulator, operator, passenger. Passing responsibility downwards to the passenger (sloping shoulders) is an easy, and commercially attractive option with minimal regulatory action; but is this ‘safe’.
Ask the passengers, but they don't want to pay higher fares; so they take the risk - ‘it wont happen to me’, or if it does ‘I will know what to do’, until the fire is too hot … … smoke …
Meaningful change is unlikely without a very public major accident; cf Manchester 737. A sad sorry state for future safety thinking in a very safe industry, but we were warned, see work by R. Amalberti ‘The paradox of ultra-safe systems’.