the absence of evidence is evidence of absence (all A is B, B therefore A logical fallacy).
This, of course, is nonsense
You are trying to equate an immensely complex probablistic logical construct (absence of evidence is evidence of absence) with a very simplistic statement of predicate logic. What we reallly have is NOT EVI(A)-> EVI(NOT A). Can you give me a definition of EVI()?
I think it is obvious that absence of evidence is evidence of absence for all "reasonable" definitions of "evidence". This may be extremely weak evidence, but it is evidence none-the-less. We are essentially performing a search of a parameter space for a particular target. Null-results simply reduce the available parameter space for our target. The confidence in a declaration of absence will depend upon the known parametric extent of the target, and the remaining available parameter space. Ever played battleships? What is the probability of finding a battleship in an unsearched 4x4 grid, that is surrounded by null results?
And whilst we may be only generating *evidence* of absence, we are gaining knowledge of the constraints of target...