Earlier Thadeosaurus (Carroll 1981, Figs. 1-3) nested close to protorosaurs in the large reptile tree (still needs to be updated), as a sister to Tangasaurus at the base of the Enaliosauria. Another look at Currie’s (1984) tracings (Fig. 3), rather than Carroll’s 1981, 1993 reconstruction (Fig. 1), inspired a new reconstruction (Fig. 2) and nested it within the Enaliosauria, on the other side of Tangasaurus, between Thadeosaurus and Acerosodontosaurus. So this new nesting shifts Thadeosaurus a few nodes. Thadeosaurus usually nests with tangasaurs. So everyone is in agreement here.
At this grade, these basal enaliosaurs show no obvious aquatic adaptations. Rather, Thadeosaurus appears to have been a long-legged sprinter. The skull remains very much like those of basal diapsids. No special features there.
Sometimes you just have to employ more than one specimen to create a chimaera. In this case, Thadeosaurus was reconstructed from several specimens of various sizes (Fig. 3). Hopefully these are all congeneric and conspecific specimens, as reported by Currie (1984). I didn’t see any red flags here. Nothing about the reconstruction is at odds with Currie’s observations.
When genera look so much alike, as basal enaliosaurs do, it is paramount to get the details right. Try to get back to the original material. If not, try to get back to the in situ tracing. If not, use the reconstruction and hope it is accurate.
Carroll RL 1981. Plesiosaur ancestors from the Upper Permian of Madagascar. Philosophical Transactions of the Royal Society London B 293: 315-383
Currie PJ 1984. Ontogenetic changes in the eosuchian reptile Thadeosaurus. Journal of Vertebrate Paleontology 4(1 ): 68-84.