S116* is far more widespread than L21 making it pretty hard to narrow down its point of origin. The problem is we do not know what part of the S116* paragroup is really an S116* i.e. ancestral clade of L21. Has anyone looked at any differences in the STRs from S116* in the different areas. I wonder if the S116* in any particular area is more similar to the upstream S116 negative clades. That might give a clue as to where the S116 mutation happened. Likewise where is the most L21-like S116*.
I think it seems very likely from distribution that we can at least narrow down L21's origin to somewhere in Europe north of the Alps but S116 in contrast appears on both sides of the Alps meaning it is very difficult to even guess which route it took west from the S116 negative/S21 negative areas to the south-east. The two main routes into Europe and when travelling east to west through Europe in prehistory consisted of the riverine one along the Danube and the sea-borne Mediterranean one. For me that is one of the most fundamental missing links in the story.
My suspicion, and it is largely speculation, is that P312/S116 first arose somewhere in the east, and then split into two branches. One branch entered Europe along the Mediterranean, the other up the Danube valley. It makes sense that L21 developed out of the Danube branch, and expanded from a location in central Europe north of the Alps.
Meanwhile I think the Mediterranean branch expanded up the Atlantic coast, while the Danube branch expanded to the north and northwest, and the two branches met up and intermingled.
However the P312* WTY project is finally starting to show some new SNPs, so before too long we may be able to break down some subclades currently hiding under the P312* umbrella. With luck this may tell us something more than we currently know about the history and distribution of R1b in Europe.