If an Flow is aborted by a user in the UI, the user would naturally expect that all jobs started by that flow would be aborted.

Currently the user has to abort the flow, then abort any jobs that where started by the flow, or are still in the queue. This is frustrating to a user, but may also prevent a new run of the flow from happening successfully as Jenkins may refuse to schedule a job if another invocation of that job is already in the queue.