SCETlib NP parameter DeltaLambda2 variation now around 0, add FO stat unc. group

[reimersa]

• The change in the rabbit_theory_helper is needed to digest the new variations of the DeltaLambda2 parameter in SCETlib, which are now around 0.0 instead of 0.125. The corresponding corrections are PR'ed to wremnants-data here: https://gitlab.cern.ch/cms-wmass/wremnants-data/-/merge_requests/108. Once those are all ready and merged, wremnants-data should be bumped.
• Adding the stat. uncs. in the FO prediction (used only for NNLOjet) to the pTmodeling uncertainty group as well, and defining their own group fo_stat to plot them conveniently on impacts.