I would like to throw in the monkey wrench of "new_object(LWO)" where said LWO also contains sub-lwo's. Currently I'm treating the LWO as just another array-like copy-by-ref type. Ran across this very issue while writing an LWO-aware version of deep_copy