Yes, maybe you want to include some scaling function so larger shareholders that should have more agency bear more liability. But then you are likely going to get games played to on paper to dilute shareholdings though shell companies. Unless big transparent voting blocks are directly needed for their purposes.
The problem with your second statement is that is seems to be unclear who the relevant people actually are most of the time and we just end up with fall guys. Especially when it is the environment dominating individual agency where in effect you could swap all the people and get the same outcome most of the time.