Clusters with many observations (~10 or more) causes significant slow down in the pipeline

[bcalden]

When multiple observations (around 10 or more) are used, there is significant slowdown during multiple stages of the pipeline.

Potential fix by reprocessing the observations individually instead of passing the list of observations.