Hi cam,
#1, #2)
Complex dependencies will need to be submitted by scripts (Python or Perl).
#3)
While there aren't "skip, stop, requeue" features per se, you may define arbitrary
complex dependencies with Qube, so it's possible to implement, for example, a
dependency that will stop a job downstream when a job upstream changes its
status. Needless to say, it's also easy to manually block (stop) a job.
#4)
The Qube GUI has a feature where you can apply a filter to only display jobs in the same "process group" or "pgrp"-- usually, all jobs in a dependency tree is put in a single pgrp.
Have a look at our Development related documentation, which is available on our support site
http://support.pipelinefx.com/ for more info on creating dependencies.
Hope that helps!
-shinya.