The tree of life and the rock of ages: are we getting better at estimating phylogeny?

In a recent paper, palaeontologist Mike Benton claimed that our ability to reconstruct accurately the tree of Life may not have improved significantly over the last 100 years. This implies that the cladistic and molecular revolutions may have promulgated as much bad "black box" science as rigorous investigation. Benton's assessment was based on the extent to which cladograms (typically constructed with reference only to distributions of character states) convey the same narrative as the geochronological ages of fossil taxa (an independent data set). Fossil record quality varies greatly between major clades, and the palaeontological dating "yardstick" may be more appropriate for some groups than others.