It's not like the 2016 article was some kind of a spy leak. They were allowed to cooperate on things that lead to this publication - because in 2016, there was

It's not like the 2016 article was some kind of a spy leak. They were allowed to cooperate on things that lead to this publication - because in 2016, there was a reason to allow it - NIH grants alone would probably be a good reason. As soon as they learned about the screwup (that likely happened somewhere in 2019) all evidence of the works related to the leak that hasn't been published has been hidden or destroyed. Thus, we can't expect any publications about that work, at least not from 2019 onwards. You're fishing for some contradiction, but there's none - publications do not suggest absence of censorship, publications only suggest the censors had no problems with it being published. And indeed, they couldn't know in 2016 that their research in 2019 would lead to a massive screwup with a global consequences. If they were an open society, there would be a constant trail of evidence leading to the moment of the screwup - communications, orders, articles, chats, etc. In a totalitarian society, all these were under tight control by default - and as soon as it turned out they are dangerous - which became clear sometime in 2019, when the Chinese government learned about the possible leak - all this information, already tightly controlled by default anyway, was disappeared or at least made inaccessible to outsiders. It is possible it still exists somewhere in the dungeons of the Great Firewall - who knows. The point is this is why information control exists - to make such things possible. In an open society, this would be completely infeasible - tracking back everything connected to a certain research program and making it as if it didn't exist. In a closed, tightly controlled society - yes, plausible.