Is there a performance benefit to Spark streaming application with multiple jobs (actions) vs one job?

Question

Given the following,

  /**
    * With only one job active at a time this processing can ensure that data for
    * all partitions must be written to A before