Consider what that method of evidence-cum-faith would mean in everyday life: if you caught a co-worker lying to you a fair amount of the time - not all the time, for sometimes he or she says things that are indeed true - but in a significant portion of instances, on occasions when his or her word could not be checked, would you automatically accept as your presumption that the uncheckable utterances were true? Not if you valued your own position.
[Akenson, Saint Saul, McGill-Queen's University Press (Montreal: 2000) p. 140.]
Akenson is in the process of arguing that the common use of Acts as a historical document in uncheckable instances is unreasonable because Acts on its own is in some cases at the very least misleading. Whatever may be the case with that, in his everyday example Akenson is somewhat confused; it has been known for a long time (since the 18th century) that not only do we often accept testimony under such conditions, in everyday life we regard people who refuse to do so as unreasonable (the 'principle of credulity' remarked upon by Reid, Campbell, and others, where 'credulity' is taken in a non-pejorative sense). And, indeed, we academics should be very thankful this is so, for scientists, historians, biblical scholars, and the like are one of the biggest sources of misinformation for non-academics. We're a major source of information, as well; but if people actually treated historians and the like according to the principles Akenson is suggesting, the only possible result would be widespread refusal to accept expert testimony; for people are mislead by purported experts, of all sorts of credentials, quite often. In fact, this approach to testimony would be unreasonable, since we best police testimony by accepting as much of it as we can, eliminating contradictions, filtering it through our own experience, etc. If you value your position, you might well be unreasonable to refuse to accept the testimony of a coworker who often lies but sometimes tells the truth, even if you are wise to do so with considerable caution.
[On the question of Acts itself, it strikes me that the question is rather unimportant; whether one uses Acts in this way will simply depend on what you are trying to do. Some methods are just bad historical methods, period, since they are of little value for any historical work one might be doing; but many methods are of use for certain sorts of ends, and the only bad historical work with them is using them for ends that should eliminate them.]