[dvfjsdhgfv]

There is a sea of difference between the Chinese paper and the other two you mentioned. The Chinese were very straightforward: can we identify facial traits that are more present in criminals? To a certain degree, they answer positively by saying there is a greater variation in facial traits in criminals than in the rest of the population. Of course there was a lot of criticism of their methodology (to which they answered [0]).

The first article you mentioned describes the bias of eyewitnesses related to stereotyped "criminal look". But the researchers are very clear: "Further research is needed to identify the features that are associated with the criminal stereotype and how they affect lineup decision processes. The specific elements associated with criminal face stereotypes have not yet been identified."

The second article deals with our interpretation of certain traits (which, in this case, are listed) and it's careful not to imply these traits are actually related to criminality. On the contrary: "such evaluations could inappropriately influence decision making in criminal identification lineups. Hence, additional research is needed to discover whether and how people can avoid making evaluations regarding criminality from a person’s facial appearance".

[0] https://arxiv.org/abs/1611.04135

[foldr]

The Chinese paper is almost certainly wrong. A simple explanation for why other people aren't publishing the same claim is that it's both inherently implausible and unsupported by any of the available evidence. You can find plenty of equally kooky and potentially controversial papers published in Western countries. E.g. https://journals.sagepub.com/doi/abs/10.1177/095679761245146... Of course we can play a game where you keep denying that any example I give is exactly parallel to the paper you cited. However, I'm sure that if you do a few Google scholar searches you'll be able to find something that satisfies whatever criteria you have in mind.