exploring possible topo. 3-body thresholds

Two new lines for the 3body topological triggers are added with a looser MVA cut than in the nominal 3-body lines.

Table overview of the MVA Cuts of the nominal 3body and the loose 3body:

Line Topology	Nominal	Loose
3-body B	0.99974	0.99934
3-body Mu	0.99969	0.99932

With the loosened MVA cuts, the 3body topological trigger is given more freedom in the selection, allowing recovering more efficiency, specifically in the hadronic modes. On the sample B2DPi on MC, an increase of roughly 15% efficiency is observed using the EfficiencyChecker.

Cc @nschulte @abertoli

• a (detailed) presentation on the topo trigger performances has been given by @mbovill at the WP3 Selections meeting on Sep. 16th, see this link
• a one sentence summary of the presentation above is that the efficiency of the 2-body topo. relative to the exclusive line, for B2D(2KPi)Pi, using 2025 data, is above 90 %. In the Bd2D(2KPiPi)Pi case, always using 2025 data, the relative efficiency of the 2-body or 3-body is 70 %
• by loosening the mva cut on the 3-body we would like to recover efficiency for D2HHH and D2HHHH
• the LHCb GB/sec to tape variation using the newly implemented input reference sample is + 5.1 %
• extra tape cost starting from the Oct. TS until the end of pp running, assuming 60 % LHC efficiency is 572 TB

Support from the B2OC conveners (@jobutter and @cchen2):

From the B2OC management side, we would like to express strong support for loosening the HLT2 topo 3-body trigger. 
We would like to stress that in B2OC, all our Sprucing lines depend on the Topo 2/3 body HLT2 lines. In the Working Group Meetings, a wide range of studies have shown that the HLT topo efficiency is lower than desirable and there is clear support to improve this. 
The changes in this MR will be very beneficial for B2OC and LHCb.

Support from DPA (@nskidmor)

DPA is happy with this decision and will accommodate any disk space mitigations should they be required. 
The FULL stream (largely the topo) is the basis of LHCb's legacy physics programme and therefore this study is crucial to optimise the physics potential of LHCb long-term. 
What we learn from this exercise will inform 2026 datataking

Support from the B&Q conveners (@elspadar and @dtou)

B&Q is also supporting this decision. B&Q is currently using Topo 2/3 body HLT2 lines for b->etac X decays. A summary of the B&Q usecase and future usage can be found here: https://codimd.web.cern.ch/MRC0VbI9R22LSbukLzQWEQ?both.