Unfortunately, it seems that it's an "overall" model -- see under "Study Design" in the "Methods" section of the paper.

There are "individual harms" that take account the harm done to the user himself or herself.

But, "the extent of individual harm is shown by the criteria listed as to others take account indirectly of the number of users"

(emphasis mine)

Unfortunately, parts of the study are well-thought, but others, not so much.