COB-85: LMM implementation and testing

This is a major update where many things have changed to get multiple beamFormer pipeline functionality working.

A number of notable changes:

• Fixed a bug in reading the quantization settings from the Parset
• Fixed TAB indexing, e.g. for TAB delays and filenames (major impact throughout RTCP)
• Fixed bug in flagging in BeamFormerPreprocessingsStep when a station subset is used
• Allow station subset in BeamFormerCoherentStep and BeamFormerIncoherentStep
• Fixed subband selection in SubbandProc, and properly record this selection in BeamFormer::File
• Added support for per-beamFormer setting of doFlysEye
• Add two new integration tests: tBeamFormerStationSubsetSubbandProcProcessSb and tBeamFormerSubbandSubsetSubbandProcProcessSb
• Miscellaneous cleanup and general improvements to earlier changes made in COB-85 commits

The new functionality is tested by running various individual parsets (with recorded data) for the LMM beamFormer submodes and comparing this to the output of a run with an all-in-one parset. This comparison is successful, indicating that the LMM multiple beamFormer functionality (more or less) works.

TODO:

• Add an integration test for the aforementioned correctness test