From a certain zoom level on in the scaffold tree view it is difficult to take advantage of
the property mapping onto visual cues, like the node background, for larger data sets.
On the other hand, on these zoom levels (e.g. full tree view) the area usage of the scaffold tree visualization is rather bad compared to space filling approaches. As due to the semantic zoom no scaffolds are shown, there is no well defined node size that has to be preserved, and thus the size could be changed to improve the visualization.
A helpful feature therefore would be to at least make full use of the available area on the
tree levels by expanding the nodes to cover the full area without overlap, similar for example to the representation in [1].