After a quick and very superficial internet check,
It is interesting to see how the studies on this period have progressed over the last decades.
As we now have a focus of rescue archaeology,
I'd like to throw in a few notes on the apparat for directed research here.
While they have their own difficulties , contemporary written documents are one of the densest stores of data,
and can provide many excavations worth of data.
And since data is distributed by writing, they are particularly useful.
You find contemporary written documents at governmental structures, religious structures,
high status residences, high status burials, and at major public works.
Secondary writings have to be very carefully analyzed to be of much use.
(min, this applies to the Bible, the Bronze Age Aegean oral cycles, and the medieval Arthurian materials.)
The best that can be hoped for is simply to indicate a location of interest for excavation.
( A very thorough understanding of ancient technologies can lead to sites located at resources or along trade paths.)
My view is that a large problem here was destruction of writings in the British Museum Library fire.
Most of the recent work for this are area and period appears to be the result of analysis of later Welsh materials:
https://en.wikipedia.org/wiki/Coel_Hen
Which is why the excavation at that ring fort is so important.
For this area and era, from my very thoroughly limited view,
there are two important writings which need to be thoroughly deconstructed:
Admonan's LIfe of Columba, and that terrible translation of the Prophatio Merlini.
Now to work with any writing, you have to know the language,
and in particular place names and titles.
You have the Pict constructions, which have their own uses.
Not much work has been done with anglo saxon title "Aed" variants,
such as Aedan mac Gabran or Aedan Artur.
And then there is the problem of the title "draign" and "dragons"
such as:
"This small enamelled bronze Roman trulla (ladle), dating to the 2nd century AD, is inscribed with a series of names of Roman forts along the western sector of the wall, together with a personal name and phrase: MAIS COGGABATA VXELODVNVM CAMBOGLANNA RIGORE VALI AELI DRACONIS."
where DRACONIS is being read as a variant of a personal name instead of as a title.
In sum, without a detailed understanding of the local situation as regards foederati and mercenaries,
you can nor analyze it in an imperial context.
You can try to use such a global analysis to throw light on the local situation,
or to limit possibilities,
but it is way down my list of methods of directed research.
Thu my opinion is that the excavators of that ring fort need to focus their excavations on the recovery of contemporary writings from the site.
I hope they have done their metal detector sweeps and their ground penetrating radar and resistivity surveys and their Lidar map.
Once again, my research focus is on impacts, in his context particularly the one at Bazas,
which should be showing up in the Irish records as well.
All of the data has to fit together, and smaller recent impacts are very useful for chronological work.