No one claimed infections "top out" at 10-20 %. I agree, that is CLEARLY untrue. Rather, they seem to "level off" at 10-20%. That is, the rate of spread/new infections stops increasing so fast. Exponential infections always follow an S curve, but usually the leveling off of new infections starts at much higher penetration levels (e.g., 50%).
I mean, perhaps I misinterpreted the original article, but it sure sounded like the author was hypothesizing that T-cells confer immunity to a large percentage of the population which is why we were only seeing infection rates of up to 10-20% in the wild. I'm not sure his theory would make much sense if he actually meant "the rate of spread stops increasing so fast after 10-20% but continues up to 60% anyway".
Your confusion seems to lie in the implicit assumption that the "immune" can never get infected. That's not how immunity works, in general. Immunity is a level of protection not an infallible shield.
