I am saying your definition of evidence is overly strict and at odds with both how we use the term in common speech and with what is useful. My definition of evidence for A is an observation B such that P(A|B)>P(A), and with this definition, lack of evidence most certainly is evidence of lack.