gstlal_kafka_dag: remove explicit accounting user/group information and use --condor-command only for consistency

update gstlal_ll_dq, gstlal_ll_inspiral_aggregator to work with API change in ligo-scald introduced in 0.2

- was using LIGO_SEGMENTS_VERSION instead of GLUE_VERSION to decide when to
  print the warning about a too-old glue version

- enable version checks for required python-ligo-lw

running offline inspiral jobs with gracedb uploads. The command line arguments
are constructed from a the process params table in an injection results
database

interpolator: fix bug where interpolator was pushing out nongap buffers when it received gap buffers (chads patch)

- bump min glue version to 2.0.0

- bump minimum required python-ligo-lw to 1.5.3 (but don't enable check for it, yet)

- add M4 code to check for a functioning minimum required version of
  python-ligo-lw.  not yet used

- gstlal-calibration is not touched

gstlal_ll_inspiral_aggregator: tweak kafka consumer settings to mitigate heartbeat session issues, close connection to consumer upon exit

gstlal_ll_inspiral_pipe: scale up number of aggregator jobs to keep up with real-time, gstlal_ll_inspiral_aggregator: add more specific group_id to not conflict with other aggregator jobs

remove gstlal_ll_inspiral_state from gstlal-ugly as gstlal_ll_inspiral_aggregator can now handle state aggregation as well

gstlal_ll_inspiral_aggregator: use new aggregation classes from datamon to aggregate data, tidy up imports

gstlal_ll_dq: use new aggregation classes from datamon to aggregate data, tidy up imports and add copyright info

- bump min ligo-segments to 1.2.0
- also set gstlal-burst's package metadata to depend on ligo-segments instead of glue-segments

aggregator execs + ll_inspiral_pipe: allow more control with number of routes processed per job, increase number of aggregator jobs in online dag for scalability