There are a two issues here:
- Do you think that the non-balanced growth is a “steady state” feature or something that happens during the transition to a balanced growth path (e.g. catch-up growth)? I do not have much experience with models of the first class, featuring things like generalized balanced growth paths. Note also that countries like the US feature some (moderate) degree of capital deepening that is very hard to model in one-sector economies, but people consider this still to be consistent with balanced growth as there is no massive trend over time.
- Are you thinking about a stochastic model or a perfect foresight/deterministic one? Stochastic models are linearly approximated around an approximation point whose dynamics are meaningful for the experiment under consideration. In that case, non-balanced growth may be hard to model, because there is no steady state around which to approximate.