use new ThroughPut opts for nightly tests
put HLT1 reco options based on @chasse's and @sponce's approach (lhcb-core/lhcb-benchmark-scripts!3 (merged)) into Brunel as qmtests,
in order to proceed getting rid of minibrunel.
The goal is to pick up these options for throughput measurements, so that nightly and throughput tests are stronger coupled..
To demonstrate that these options work just as well as minibrunel, I used old refs for two of the tests (the ones with fit). They will of course fail (@cattanem) due to different outputs in avalanche scheduler and the new one, but the counters match and it looks consistent.
The other two tests have working refs..
This is an intermediate step (and therefore yet another hack).
There is still ongoing work to make the python configuration more flexible and nice, but there is no clear timescale for this, so we converged towards using this very thin configuration for the moment (does only rely on application manager, not on brunel/tracksys/minibrunel)
thanks @chasse for the nice name