No, Marcgrave is pre-Linnaean (he died in 1648), thus did not use "scientific names". "CARACARA Brasiliensibus" is actually a simple statement that the bird he was describing there was called Caracara by native Brazilians.
http://gdz.sub.uni-goettingen.de/dms/load/img/?IDDOC=234575 , page "355: 211".
Note that most of his descriptions begin with "[something] Brasiliensibus"...
Buffon is not pre-Linnaean, but he did not use binominal nomenclature, hence his names are not available.
"Caracara, Buff.", as such, in addition to being an unavailable name, is also in my opinion an unidentifiable
concept, because Buffon used this name for two completely different birds in his writings. One was indeed Marcgrave's bird, that is now deemed unidentifiable. The other (see last link in my previous post) was a bird described by Reverend Father Jean-Baptiste du Tertre--a capon-sized bird observed in the French Antilles, called "faisan" ("pheasant") by local French people, longer-legged and longer-necked than a fowl, with a blue neck and breast, a grey back, and short black wings and tail. Buffon named this bird based on its call, described by du Tertre as "caracara-caracara". Du Tertre also made clear that this bird used to be domesticated and, when so, behaved very dominantly, and that its flesh was tasty. This bird has been thought to be a trumpeter by several subsequent authors (including Buffon himself, at a later stage), but there are no trumpeters in the Antilles. Based on the description, I actually think feral guineafowls would be a much better guess (these were introduced in the Antilles back in the 17th C).
Anyway, my concern in the present case is due to:
67.2. Species eligible for type fixation (originally included nominal species). A nominal species is only eligible to be fixed as the type species of a nominal genus or subgenus if it is an originally included nominal species.
67.2.1. In the meaning of the Code the "originally included nominal species" comprise only those included in the newly established nominal genus or subgenus, having been cited in the original publication by an available name (including citation by an incorrect spelling [Art. 67.6]) of a species or subspecies (see Articles 45.6 and 68.2), or having been cited there as the deliberate application of a previous misidentification (see Articles 11.10, 67.13 and 69.2.4).
From this it seems to follow that "Caracara, Buff.", the only species cited by Vieillot when he introduced
Polyborus in his
Analyse d'une Nouvelle Ornithologie élémentaire, is not eligible to be the type species of the genus. The genus probably fulfills:
67.2.2. If a nominal genus or subgenus was established before 1931 (in the case of an ichnotaxon, before 2000 [Art. 66.1]) without included nominal species [Art. 12], the nominal species that were first subsequently and expressly included in it are deemed to be the only originally included nominal species.
All of the discussions I have read about this case have focused on the (un)identifiability of Marcgrave's bird, that forms the base of (one of) Buffon's Caracara(s), itself assumed to be the type species of
Polyborus by monotypy.
All of these discussions become irrelevant if Buffon's Caracara cannot be the type of
Polyborus.