How do you come up with your assumptions ?
"Assuming that the large majority of PCS are resolved after 3 years (in the 4th year) and assuming that the average yearly number of concussions are 2 million (and assuming no repeat concussions for individuals) I get an estimate of the number of people currently dealing with prolonged PCS in the United States: 54000
This amounts to 0.02% of the population. This means this is a rare disease. "

What I've read is 15% of concussions have symptoms past 6 weeks. These are the classic definition of PCS. The average annual number of concussions is just those that show up in ER's or Urgent Care. Estimates suggest these 1.5 to 3 million is about a third of all concussions.

From what I've read about long term or even never resolved symptoms, the distribution is a version of a bell curve with a long tail to the right. As one study said, The longer the recovery takes, the greater the chances of a very long recovery. Ignore the left half of the bell curve. Notice how the right end starts to flatten out. A real curve of concussion recovery would be high and steep in the middle and much longer to the right with a long slow slope to the right.