Bring changes from nh MQ online selection in general form to version with MUCH + various fixes

Of interest to @a.toia @v.singhal @n.herrmann

1. Option to bypass the consecutive TS buffering + re-ordering in Sink
   => [Tested with run 2160, 1 tsa file (50% missing TS), 4 branches and 30 TS]
2. Selection of events with a minimum number of Fired detectors
   => [Tested with run 2160, 2 tsa file (no missing TS), 1 branch and 30 TS]
   => Agreement of distributions within 10-20% with same conditions for NH MQ selector, difference seems to be because
   • in my case the T0 is not adding fired TOF detector counts, leading to ~1500/42000 events being rejected
   • the parameter files and setup files cannot be made exactly the same in the 2 MQ scripts
3. Fix the MUCH parameter file and file loading for runs before 2162
4. Improve/fix the handling of the error objects in the STS unpacker
5. Fix the BMon mapping in both MQ, task and monitoring uses
6. Prepare the TOF unpacking files to have 1 parameter per March 2022 beam on target period

Not sure if we should wait for further review or just include it in the mCBM master, as most people will be booked for the "main CBM @ SIS18" investigations. If ok I would merge it tomorrow morning, otherwise next week but that would probably delay the merged from mCBM master to official master.

Not included due to lack of time for coding it but found during the preparation:

• we could probably speed up further the event building by checking the trigger conditions of a given detector system immediately after selecting its data instead of after all systems are through.
• This would be especially efficient in the case of selection on min fired det nb in the reference detector as then most noise event seeds would be skipped before looping on the data of the other detectors.