Interesting. However, your sources are rather obscure, and I'm not sure what you mean by "a weak form."
I'd say a Baysian analysis is a way to evaluate the veracity of an argument from ignorance, not necessarily a replacement of that argument. An argument from ignorance is a strict subset of a baysian analysis in which there is no causal relationship between the missing evidence, and the conclusion. This is such a case. So we can conclude by using a baysian analysis that this is in fact an argument of ignorance.
Unless, of course, you have additional statistical samples that show a non zero relationship between not knowing about NSAs activity and the fact that they are engaged in that activity.
http://en.wikipedia.org/wiki/Argument_from_ignorance