Oh - and there's signs that infection starts off in the mouth (and maybe respiratory tract) (https://unglossed.substack.com/p/travels-to-maskladesh#footnote-anchor-5), which possibly accounts for the fact that PCR testing wasn't such a bad correlate for natural infection as was believed - you can't scrape coronavirus off the bottom of someone's brain-pan until there's at least an amount of asymptomatic spread that correlates with a later memory immune response. Agree that it's odd to be surprised to find antibodies in saliva.