Reconstruction of transmission chains (i.e. who infected whom) is a sophisticated subfield in infection disease epidemiology. This atrocious drivel masquerading as science, with no stats, no genomic sequencing, no nothing ... besides a childish diagram is not sophisticated.
1/

29 people got infected, including 8 children, that's it. The authors conclude "The high infection attack rate among children in our cluster *could* be explained by prolonged close contact between very young children, who are less able to adjust to control measures."
2/

Indeed, children *could* have infected other children/adults, or not, the data doesn't allow to tell. Pretending otherwise is just speculation. The only saving grace of the paper, is that the authors are at least upfront about having failed to collect meaningful information.
3/

Unsurprisingly a hysterical nutritionist with a large Twitter following had a field day reporting it and coming up with further interesting details in the process.
4/