Woah are we actually debating this again? Baseops crashed for 48 hours and I thought we all went through a day where we felt lost and helpless, followed by another day where we reconnected with our families and remembered there was a life outside of BO.net and arguing on the internet.
I've posted this before but 19andMe has done one of the leading calculators for estimating cumulated risk of contracting COVID-19 based on a meta analysis of dozens of studies followed by a pretty impressive mathematical model.
I ran the model for a 26 year old slightly healthier than average male in my zip code, unvaccinated, but taking basic social distance precautions based on the current local measures we had in place. This represents the average US Armed Forces enlisted service member.
Since the calculator only predicts risk in a 24 hour window, I had to accumulate that risk over a period of 1,095 event occurences to build a 3 year risk model.
https://19andme.covid19.mathematica.org/
To your other point, I do not believe the occurrence rate will approach 100% as that is not what historically happens with viruses. There is a mathematical model that explains the waves and valleys of virus transmission and even during some of the most infectious pandemics in human history, the vast majority went without ever being exposed.
Edit: rereading your post your method is flawed for several reasons. 1.) If the CDC estimate is really so far under you would then need to extrapolate that reduced risk to your remaining risk because people who are not reporting COVID-19 are not doing so because they do not have symptoms, hence, no problems. 2.) I don't think your risk is correct. If we just estimate 44% is ~1/2, bro, I've only known of like 6 people with COVID the whole pandemic. That's my whole base population since anytime there is a positive case here we shut down literally half the base and it's rather obvious. Do you really know that many people dropping off left and right that this passed your common sense test?