Yes - it is. In my code, just above the snippet I originally posted, I have the following:

int RT_sub_brick_id[2];

RT_sub_brick_id[0] = 1;
RT_sub_brick_id[1] = RT_written;

so that should take care of the requirements of THD_copy_dset_subs ... I will rework things a little, try to correct the "incoming data" and "write" conditions for the original (multichannel) and volume registered data - which are not identical, though I have put them as such in my initial test code ...

Just a general question about dataset creation and writing - I copied the single time point sub-brick into a new data set. Does THD_copy_dset_subs create the new dataset ? I thought that's what the EDIT_empty_copy function did ...

And after that, I edit the name of the data set before it's written out to disk. Is there anything else that needs to be done before one can write out a dataset to disk ? Even when that data set was copied from another ?