Historians can't prove anything. No social scientist can. Proofs require data collected through experimentation in a controlled environment -- and I can guarantee we won't be letting historians invade Guatemala to test out their theories any time soon.
Historians can certainly prove things, just as any reasoner does. Strictly (e.g. logically) speaking, a proof is a finite sequence of statements each of which is either a premise or else follows from the preceding statements by the application of an inference rule. There is nothing peculiar about historians or social scientists which prevents them from employing logical reasoning in such a way. (Besides, no empirical notion such as ‘experimentation in a controlled environment’ enters in the standard definition of ‘proof’

. Perhaps, unlike, say, physical scientists, historians are much less likely to produce anything resembling a mathematical proof, but not all valid proofs (even strictly formal ones) need to be based on a mathematical model of the empirical data.
Furthermore, you've taken a necessary element of the Spanish conquest of the new world (native support) and turned it into a sufficient one. Think about it like this: had the Tlaxcalans, along with other native peoples, been capable of defeating the Aztecs without Spanish support they would have done so long before Cortez landed.
In fact, the condition stated by Perceval may be fairly plausibly taken to be a sufficient one, as follows:
(1) If (A) the attackers are supported by the natives , then (B) they will defeat the Aztecs.
Thus stated, (A) is indeed a sufficient condition for (B). In order to deny that, one has to show that (A) doesn’t imply (B)—which, I think, is what you have tried to do with your Tlaxcalans counter-example: if the native attackers had been supported by none but themselves, and not by outsiders, then the Aztecs wouldn’t have been defeated. However, even if we concede this point (but let me notice in passing that your example seems off the mark in our context, for the supported attackers are supposed to be outsiders, and not themselves natives) it is not at all clear that Perceval has committed the fallacy of mistaking a necessary condition for a sufficient one. To claim that (A) is a necessary condition for (B) – which is what you appear to be saying – is to claim that (B) implies (A), e.g.:
(2) If (B) the Aztecs have been defeated, then (A) the attackers were supported by the natives.
But, of course, to show that (2) is the case is by no means a refutation of (1), for (1) and (2) may be simultaneously true. In other words, taking a necessary condition to be a sufficient one need not be a fallacy, for a condition may be at the same time both necessary a sufficient, in which case we have an equivalence.