Agree, this paper has a series of figures that lead to a 3d representation of 3 axes I work out, so this builds up logically and it is not about interpreting the purely example dataset, but getting the idea of clustering in 3dimensional space. So I think the figures should be fine here.