That would be a viable argument if everybody killable has already been infected.
If everybody killable has not already been infected, then the bend in the curve would mean that something has apparently slowed their rate at which they are being infected.
If some portion of the overall population has already been infected, lets say 10%, that is a stochastic process: It will be 10% of the vulnerable population also. Viruses don't go from door to door and only infect people they are going to end up killing.
Althought it is always fun to know the conclusion first then try to fill in some kind of reasoning afterwards :-)