Here’s the abstract of a recent paper that argues that the Dmanisi hominids of Georgia are closer to Homo erectus than Homo habilis:
Anatomical descriptions, comparative studies and evolutionary significance of the hominin skulls from Dmanisi, Republic of Georgia. G. Philip Rightmire, David Lordkipanidze, & Abesalom Vekua. Evidence for ancient hominin occupation in Eurasia comes from Dmanisi in the Georgian Caucasus. Stratigraphic and sedimentological arguments, geochemical observations, paleomagnetic sampling and radiometric dates all point to the conclusion that bones and artifacts were deposited at this site during a brief interval following the close of the Olduvai Subchron (1.77 million years ago). In this report we present further descriptive and comparative studies of the D2280 braincase, the D2282 partial cranium, now linked with the D211 mandible, and the skull D2700/D2735. The crania have capacities ranging from 600 cm3 to 775 cm3. Supraorbital tori and other vault superstructures are only moderately developed. The braincase is expanded laterally in the mastoid region, but the occiput is rounded. The pattern of sagittal keeling is distinctive. D2700 displays a prominent midfacial profile and has a very short nasoalveolar clivus. Also, the M3 crowns are reduced in size. Although there is variation probably related to growth status and sex dimorphism, it is appropriate to group the Dmanisi hominins together. With the possible exception of the large D2600 mandible, the individuals are sampled from one paleodeme. This population resembles Homo habilis in brain volume and some aspects of craniofacial morphology, but many of these features can be interpreted as symplesiomorphies. Other discrete characters and measurements suggest that the Dmanisi skulls are best placed with H. erectus. There are numerous similarities to individuals from the Turkana Basin in Kenya, but a few features link Dmanisi to Sangiran in Java. Some traits expressed in the Dmanisi assemblage appear to be unique. Reconstructing the evolutionary relationships of these ancient populations of Africa and Eurasia is difficult, as the record is quite patchy, and determination of character polarities is not straightforward. Nevertheless, the evidence from anatomical analysis and measurements supports the hypothesis that Dmanisi is close to the stem from which H. erectus evolved.
The link came from the Palanth forum.