Some information on how to read this table:
Time Remaining - this is the number of seconds before a scheduled job is expected to run again ( ie. it is the Next Start - Current Date/Time )
OverDue - this is the number of seconds that a job has been waiting to execute once it has been triggered to run. Please note that if multiple jobs are scheduled to execute at, or near, the same time, only 1 thread is spawned ( this is configurable ) and the jobs are executed in series. If one of the jobs takes an exceptionally long time to complete - the other jobs may be waiting ( and overdue ) for quite some time.
STARTED_FROM_BEGIN_REQUEST - this means that you are using Request mode rather than Timer mode. In Request mode, it means that if your site is not getting much traffic, the scheduled jobs will not trigger on their expected schedule and you may end up with many overdue jobs executing ( and you will likely see many jobs with the same Next Start date/time ).
With all the configuration options available in the Scheduler, it is next to impossible to solve these types of issues through Forum posts. It really requires direct access to the system to perform analysis and diagnosis in the operating environment. This is what we do through our Support offering, and we have helped many users to optimize their scheduled jobs in the past.
I hope the information I have provided has been helpful...