The algorithm seems rather easy to implement (a couple of self-contained lines) but only produce the direction field. But hidden behind this simplicity are less trivial tricks (e.g., the paper mentions that the choice of the next free simplex is guided by a harmonic field, which requires solving a linear system ; the field is further processed by a Laplacian smoothing operator...), which is why I didn't rate the algorithm higher in simplicity.
If you want to contribute with another review, please follow these instructions.
Please consider to cut/paste/edit the raw JSON data attached to this paper.