Phonetic similarities are quite the "historical" hit these days and the funny thing is that they're almost always done by non-specialists (our most famous example is Petar Dobrev, a professor in economical history, the greatest propagator of the East-Iranic theory for the Bulgars, using words from high and low without any *scientifical* linguistic connection). One such example is
this map. Funny thing though - f.e. it shows a place called Varnu in Afghanistan (or Iran, not sure) and our Varna in Bulgaria; but while reading the Heimskringla I found a direct Varna in Sweden as well (obviously, we have no connection with Sweden, so the Varna-Varna phonetic "similarity" is void, as can be in those other cases). True, if there's a whole cluster of such evidences, supported by non-linguistic (i.e. archaeological etc) sources, it could become a thesis worthy of further research, so it could then become a theory, if actual substantial evidences are found.