Well there is also two other options I can see, although I cannot say how realistic/likely these are: either the researchers „guessed“ wrong and the number in the url is no unix timestamp but just some random digits. Very unlikely imho, given how close the timestamp values are, but not impossible. The other option would be that the pcr test was indeed taken on the 16. but only entered into the system 10 days later on the 26. not sure if that is realistic. Not saying it’s either of these, just that we cannot disregard these options imho.