Epidemics don't follow exponential functions. You are aware of that?

They do until they hit their inflection point.

Which they do by themselves. Not everyone who is infected dies, survivors develop a certain level of immunity (generally speaking), cured people don't infect other people. Some people are "immune" right from the start. So epidemics slow down, the part of a given population not getting it or infecting others increases with time. Which automatically slows down the spreading of the disease.

All measures we take optimize for that. Vaccination increase the amount of immune people, quarantine pulls the inflection point forward and slows the spreading. Protecting the most at risk population further reduces the risk of spreading. And so on. So yes, the early phases are exponential. But the later phases are not. So modelling this thing exponentially against the max. number people on earth is just wrong.

All correct, but a highly contagious virus might still infect 20-70% of the entire population. Not 100% of the population, but the difference does not matter.