A simple question about a complicated object

Let n and k be positive integers with and k < n. Then of course SU(k,1) is contained into SU(n,1). Moreover, which is less clear - but proved by Khoroshkin -, the representation theory of SU(k,1) at the generalized infinitesimal character of the trivial module can be fully (and even Ext-fully) embedded into that of SU(n,1). Here is the obvious bet: This embedding is implemented by the cohomological induction functor. I conjecture that a similar phenomenon occurs whenever SU(k,1) is a Levi factor of a theta stable parabolic subalgebra of a reductive group.