This interesting question has two dimensions: (1) the meaning of παρέστησεν ἑαυτὸν ... ἐν πολλοῖς τεκμηρίοις [parestēsen heauton ... en pollois tekmēriois = "he presented himself ... by many tekmēriois"]; and (2) its history of translation in English versions.
The Meaning of πολλοῖς τεκμηρίοις
The key term here is τεκμήριον which, as noted in an earlier comment has a technical sense in rhetorical contexts, as Liddell-Scott-Jones have it in their third meaning:
3. in the Logic of Aristotle, demonstrative proof, opp. to the fallible σημεῖον and εἰκός,...
The broader sense under which this is given as a sub-heading is as follows:
II. proof (properly of an argumentative kind, opp. direct evidence,...)
And this is where the Greek semantics begin to bear on the tradition of English translation of Acts 1:3. τεκμήριον is also used in the LXX (e.g., 3 Maccabees 3:24; Wisdom of Solomon 5:11; 19:13) with the sense of "tokens" or "signs" in the less technical sense. So the question is whether, in Acts 1:3, it is this more general sense that is in mind, or the more technical, restricted sense.
There has been investigation of this diction. In older commentaries, it was sometimes suggested that this locution was at home in "Doctor Luke's" medical vocabulary, e.g. R.J. Knowling in volume 2 of the Expositor's Greek Testament (London, 1897), p. 52:
Although in a familiar passage, Wisdom v. 11, τεκμήριον and σημεῖον are used as practically synonymous, yet there is no doubt that they were technically distinguished, e.g., Arist., Rhet., i., 2.... This technical distinction, it may be observed, was strictly maintained by medical men, although St. Luke may no doubt have met the word elsewhere. Thus it is used by Josephus several times, ... it is also used by Thucydides, ii., 39, to say nothing of other classical writers. Galen ... states that rhetoricians as well as physicians had examined the distinction; Hobart, Medical Language of St. Luke, p. 184.
In a brief but widely-cited article, David Mealand examined this phrase and its usage across Hellenistic Greek literature, as its use in Acts 1:3 had struck some earlier commentators as unusual.1 Mealand demonstrates that the παρίστημι (in context, "presented") + τεκμήριον combination is "normal in Hellenistic Greek", but especially at home among the historians (he cites examples). This would make good sense in explaining Luke's usage in the "prologue" to Acts -- moreso than him being a medical man, I would think, although the "sure symptom" sense noted in LSJ has some appeal.
English Translation Tradition
If one believes that τεκμήριον is being used in its more rhetorical, restricted sense, then one might wish to indicate that by suitably qualifying "proofs" to show that this is "knock-down argument" stuff, by using "convincing" or "infallible" as a modifier. This sense entered the English Bible translation tradition with the Acts 1:3 rendering of the Geneva Bible of 1560 (fifty years before the KJV, of course):

In the 1599 edition, the notes changed, and this precise rendering attracted a comment:

The issue here is that the "proofs presented" are not of the logical kind, as required by the more "technical" sense of τεκμήριον, since the post-resurrection appearances and actions which seem to constitute the τεκμήρια "presented" by Jesus are not of this kind (i.e., arguments). However, even with the "medical" sense in mind, it is fair enough to render the Greek with a modifier to indicate this force -- if one believes that it is in this sense that the term is being used.
Clearly, it was this decision which accounts for the Geneva Bible's rendering, and which then influenced the KJV translation tradition.2
Instead of a Conclusion
I don't know why particular English translations abandoned the modifier, although both the English and American revised versions (1881 and 1901, respectively) use the simple "many proofs". Perhaps the term "infallible" brought its own problems, given the dogmatic associations of the term -- I wonder if "convincing" had been used whether it might have stuck.
Notes
- David L. Mealand, "The phrase 'many proofs' in Acts 1:3 and in Hellenistic writers", Zeitschrift für die neutestamentliche Wissenschaft 80.1-2 (1989), 134-135.
- I find it interesting in this context to note that Calvin discusses this precise nuance, implying that these "proofs" are not "arguments" (contra Erasmus), and refraining from using a modifier. Also interesting that Luther's 1545 translation did not give this sense, with its "mancherlei Erweisungen" = "various proofs".