After using the compound for a real worldish scenario, I have found a few things that would improve the workflow and be advantageous as more compounds are added.
On the Base, somebody asked for a DNA helix, and it was very simple to put together. However, some improvements I can think of:
-> Detach the Set Particle Data into its own compound
-> Specify a "Particle Type ID" attribute for every cloner compound
-> Together with the detached Set Particle Data, one could then control particle data per cloner without having to use separate ICE trees*
-> If the red compound problem can't be fixed, also detach the Geometry Array into its own cloner compound
-> Having a few more, but ultimately smaller compounds is probably preferrable anyway
*alternatively, one could specify the particle data in the Add Point node upon creation, but this does not allow users to later change particle data per cloner