I can answer this.
The x-ray appearances of the pneumonia associated with the novel corona virus were unusual, the CT scan appearances of the lungs pretty well unique. No other infection causes a similar pattern of changes in the lungs. The marked hypoxaemia without breathlessness is also unusual. The blood tests point towards an atypical pneumonia, probably viral.
Regardless of tests for virus, the medical doctors looking after these patients would have recognised that these cases were different from other cases of pneumonia, based on the symptoms, blood tests and imaging. Then add on that tests for the usual suspects were negative.
Then add on the postmortem findings that showed an unusual pattern of damage not typical for other causes of pneumonia, and the finding on electron microscopy of coronaviruses in the lungs.
In summary clinically there are enough unique features that one can diagnose covid-19 pneumonia reliably without needing a PCR test. There are enough unique features that the doctors treating the patients would have had a high suspicion they were dealing with a novel infection.
Cluster has a technical meaning, it means a group of cases linked by a common factor. So you might have a cluster of diarrhoea in a school following an outing to a farm. The worry is this might be an E. coli O157 outbreak, but at the beginning it is just a cluster of children with runny tummies after a school outing. If on testing there is no common infection the it remains a cluster. If there is a common cause it becomes an outbreak.