Carpolestes has a thumb, but that doesn’t make it a pre-primate.

In their paper on primate origins, Bloch and Boyer 2006 report,
“Extant primates are distinct from other eutherian mammals in having large brains, enhanced vision brought about in part by optical convergence, the ability to leap, nails on at least the first toes, and grasping hands and feet.”

Well,
maybe not so distinct, at least in this list of traits, after phylogenetic analysis…

In the LRT
the basalmost placental, Monodelphis domestics (which does not have a pouch), the forward facing eyes have a degree of optical convergence, adults pounce on their prey and they hold down with their forefeet, which, alas, do not appear to have nails. But 3 out of 4 is coming along way…

And here’s the little cutie in vivo (Fig. 3).

Bloch and Boyer 2006 report,
“Plesiadapiformes has long been considered an archaic radiation of primates. Evidence in favor of a plesiadapiform-euprimate link was based largely on dental and postcranial similarities, whereas the absence of a postorbital bar and other cranial features in certain plesiadapiforms provided evidence against this hypothesis. An alternative hypothesis, that Plesiadapiformes share a closer relationship to extant flying lemurs (Dermoptera) than to Euprimates, has been strongly challenged and is not followed here.”

Unfortunately, Bloch and Boyer
do not consider the third alternative, the one recovered in the LRT with a plesiadapiform relationship to rabbits, not primates. Like little kangaroos, rabbits also originated in the trees.

The large reptile tree
(LRT) nests two plesiadapiformes, Carpolestes (Fig. 1) and Plesiadapis (Fig. 2), with rabbits, like Gomphos (Fig. 3), within the clade Glires, far from primates like Nothrarctus, which developed nails by convergence. When rabbits left the trees, they lost their hallux and pollex.

In the LRT,
primates arise from a sister to the tree shrews Ptilocercus and Maotherium near the base of the Triassic/Jurassic placental radiation. By contrast, plesiadapidformes arise from a sister to Henkelotherium and the more rabbit-like tree shrew, Tupaia. These taxa share long, procumbent lower incisors and a long diastema. Tupaia has a circumorbital ring, but other Glires do not. 

The Bloch and Boyer published cladogram
included only 9 taxa and no rabbits.

Bloch and Boyer report,
“Several adaptive scenarios have been proposed to explain these [primate] specializations: (i) “grasp-leaping” locomotion (3), which predicts simultaneous evolution of grasping and leaping; (ii) visually directed predation (4), which predicts simultaneous evolution of forward-facing orbits and grasping; and (iii) terminal branch feeding on nectar and flowers, which allows that grasping evolved independently of other traits. The lack of well-preserved skulls and skeletons of the earliest primates has precluded testing of these hypotheses.”

Bloch and Boyer are overlooking
the living breathing, leaping and grasping basal primate, Ptilocercus (Fig. 4), or the basalmost placental, Monodelphis (Figs. 1,2) have most if not all of these morphological and behavioral traits. So… thumbs are primitive! Derived taxa, like whales and hoofed ungulates, tend to lose them.

References
Bloch JI and Boyer DM 2006. Grasping primate origins. Science 298:1606-1610.