I think this situation is too general to draw many conclusions, but here goes...
If $A$ is closed under Cook-like reductions, then (2) would imply $B \subseteq A$, contradicting (1), so it tells you that $A$ is not closed under Cook-like reductions.
One way to paraphrase the original statement is: to make the classes equal in power requires more than a single query.
One can then ask about other intermediate types of reductions to get a better sense of just how many queries are needed and in what way e.g. are the classes equivalent under truth-table (nonadaptive) reductions? What's the best bound we can put on the number of queries ($2$, $O(1)$, $O(\log n)$?) Since these are counting classes, one could also ask about parsimonious reductions.