Doolittle, W.F. (1999) Phylogenetic classification and the universal tree. Science, 284, 2124-2128. [link]
Interesting review discussing recent findings which question a strict tree model for the universal tree of life. Lateral gene transfer is non-trivial, especial in archaeal and bacterial genomes. Doesn't dismiss the usefulness of molecular phylogenetics as a tool, but questions it as an end-goal (producing classifications).
If there were believable genealogies of all genes… one could then ask which genes have travelled together for how long in which genomes, without an obligation to marshal these data in the defense of one or another grander phylogenetic scheme for organisms.
Nifty figures, also.