Which makes 24 days an extreme outlier

Median and range really says fairly little about the distribution.

Well yes. Yet, since there are so many unknown variables it does not seem incorrect to assume the maximum entropy distribution - Gamma.

It's already an outlier event, it's worth considering the worst cases w.r.t. asking where people have been to determine exposure.

Sure, edited out. Thanks

The most recent study I could find is here:

https://www.medrxiv.org/content/10.1101/2020.02.28.20028068v...

"The mean incubation was 8.42(95% confidence interval [CI], 6.55-10.29) days... COVID-19 course was approximately 2 weeks."

Fairly small sample size of 55 patients, which means the reliability may be debatable, and giving figures to two decimal places is meaningless. (In fact any decimal places are meaningless, unless someone times the exact moment of exposure.)

That aside - it seems very unlikely that 24 day incubations will be common.

Broadly it seems most people who become symptomatic will show symptoms within 10 days and will recover within two weeks after the symptoms appear. A small percentage - mostly older and unwell - will become seriously ill, and an even smaller percentage will die.

Meanwhile the presence of untraceable infections strongly suggests that a significant number of people - possibly a majority - either don't develop symptoms at all, or don't consider them serious enough to require medical attention.