Finalise workflow of scale and smearing correction / systematics

We need to include a proper application of correction and variation for s+s.

As confirmed with Rajdeep over e-mail, the up/down variations for scale are applied to MC. However, the nominal correction is only ever applied to data and never to MC, while the variations are never applied to data. I quote: "There are entries in the files with run-range 1-2 (thus getting picked up when MC is run), with the central value of the correction as unity, but the uncertainty corresponding to the relevant single ele/pho category. This way when run with systematics on you have the means to propagate the uncertainty, but the central correction on MC objects is still unity." So we should make sure that, when we run with correction scale, the nominal value is applied to the data samples. However, when we want to have the variation branches, no nominal correction is supposed to be applied. It is fine to read the nanoAOD run branch as stated above when retrieving the variation values.

For smearing, the workflow is simpler because everything is applied to MC.

c.f. https://gitlab.cern.ch/HiggsDNA-project/HiggsDNA/-/blob/master/higgs_dna/systematics/photon_systematics.py c.f. https://gitlab.cern.ch/HiggsDNA-project/HiggsDNA/-/blob/master/higgs_dna/workflows/base.py#L359 (we should attack here to have a consistent implementation of this rather complicated scale correction+variation business).