Haider asks if CMS correlates the type A and type B uncertainties. Ans: no (at the moment, can be changed).
Kerstin asks a technical question about whether we re-derive the type A and type B uncertainties for each differences signal parameterisation. Ans: no, the only thing that changes between parameterisations is which nuisances are included in the fit.
Some general discussion about the dashed boundaries, their purpose, and the differences of using them wrt the current CMS approach. It is commented that the ATLAS and CMS approaches are for the most part very similar.