Why doesn’t my Claude coworker run these agents in parallel? Also, what tips are there for me to speed up my Cowork processing time and improve efficiency

This is a screenshot from the background tasks from analysis i was running. It ran the eight analyzation agents in parallel, but these verification agents are only running one by one. It’s allocated three different viewpoints for verification on each idea. Why can’t each idea’s verification agents, run independently and in parallel? Like 36 recompute verification agents 36 significance and 36 interpretation? Even if there’s a limit on how many you can run simultaneously shouldn’t it at least try to run the max amount in parallel instead of going one by one by one by one? Also, is there a plug-in or a skill or something that makes this financial analyzing task, faster, and more in depth and more efficient?