The following problem appears in Aaronson's list Ten Semi-Grand Challenges for Quantum Computing Theory.
Is $\mathsf{BQP}=\mathsf{BPP}^{\mathsf{BQNC}}$ In other words, can the "quantum" part of any quantum algorithm be compressed to $\mathrm{polylog}(n)$ depth, provided we're willing to do polynomial-time classical postprocessing? (This is known to be true for Shor's algorithm.) If so, building a general-purpose quantum computer would be much easier than is generally believed! Incidentally, it's not hard to give an oracle separation between $\mathsf{BQP}$ and $\mathsf{BPP}^{\mathsf{BQNC}}$, but the question is whether there's any concrete function "instantiating" such an oracle.
It has been conjectured by Jozsa that the answer to the question is yes in the ''measurement-based model of quantum computation": where local measurements, adaptive local gates and efficient classical post-processing are allowed. See also this related post.
Question. I would like to know about the currently-known oracular separations between this classes (or, at least, the oracle separation to which Aaronson is referring to).