Different common envelope stripping depending on the phase, where merger happens

Currently, we have one treatment to account for partial stripping during a CE merger. This is based on the alpha-lambda shrinkage by stripping from the stellar profile.

We still need to update how to deal with the ejected mass for the two cases:

• merging during the alpha-lambda part (that's what's currently always used); here we should get it more self-consistent to use not only the core mass, but the full remaining mass to determine the released orbital energy.
• merging during the post CEE mass loss (e.g. stable MT); this should be more like we treat other cases of getting a merger during stable MT (I think we don't have this at the moment, because we avoid those case by going to CEE first on L2 overflow); in principle all the mass during the first phase should be lost (currently we may lose less in some cases) and there needs to be a treatment on the profile doing further stripping with orbital change according to the type of the second phase.