Multipliers are currently written just to duplicate the transects. So the variance and degrees of freedom will likely be incorrect. See this chunk of code for the current implementation.
I think ideally this would involve modifying dht to be more amenable to this type of analysis.