Hi Adam, interesting stuff and just thinking aloud - is it possible the Form used in the table is partially phonetic based on how the scribes heard them at the time of writing?
Oh, one of those deceptively simple questions.
Yes, it is, though for a variety of reasons, it's less likely in the case of place names than
personal names.
Obviously all the sources we have are written and most of those will have originated locally, in the form of petitions or local administrators, secular or religious, many of whom would be Welsh speaking (to an extent, at least - it's usually fairly obvious when a clerk who'd never heard the language turned up or was doing that, but that's bitter experience of wrangling medieval handwriting). In other words most of these spellings will have come from people who knew those places.
One thing to note about that table is that we are also looking at transcription conventions which are derived from transcribing Latin. So many of those 'u's should be 'v's because in most medieval hands they look the same. So do 'n's. In Latin it doesn't much matter, in English it does as it does in Welsh. So William Camden (a Londoner) whose attempt is perhaps most off beam was probably looking at someone's handwriting with a bit of guesswork.
Personal names are a bit different because it's much more likely the clerk will just write what he (as it usually was) hears. My favourite examples include four chaps called Llywelyn (other choices are available on that one even in modern orthography) where the poor clerk was simply guessing and went with Thlewelin/Luellin and others. Some of the French attempts at Welsh names from 15th century Normandy (lists of soldiers names from garrisons there formed a large part of a project I was involved in) are decidedly eccentric but again, the clerk could only write down what he heard.
That's a long way of saying 'yes, but not especially', isn't it?
Adam