The literature on bibliometric indices for assessing scholarly impact, in particular the "h" index (Hirsch, 2005) and its many variants, is extensive, but nevertheless Ruscio and colleagues (this issue) succeed in making a valuable contribution. They have made the effort of collecting publication and citation data for no less than 1,750 researchers, thereby, allowing for a much larger empirical analysis than what is usually found in the literature. Despite its limitations, the validity analysis reported by Ruscio and colleagues provides substantial support for preferring the "h" index and some of its variants over a number of more conventional indices, like the total-citation-count index. In this respect, the paper by Ruscio and colleagues considerably strengthens existing empirical evidence in favor of the "h" index, such as the evidence provided by Hirsch (2007). In the authors' view, however, some important limitations of the "h" index, and also of the various variants of this index, remain unmentioned in the paper by Ruscio and colleagues. In this commentary, the focus is on the following three limitations: (1) The "h" index behaves counterintuitively in certain situations; (2) The "h" index does not necessarily make the trade-off between "quantity" and "quality" in the way one desires; and (3) The "h" index represents a single-dimensional approach to bibliometric research evaluation, while in many cases a multidimensional approach would be preferable.