This is … a wheel. Or, what goes around, comes around. I have discovered that in 1904 WH Stevenson doubted the identification of arx Cynuit with Countisbury. However, Hill & Rumble, The Defence of Wessex: The Burghal Hidage and Anglo-Saxon Fortifications, 1996, p. 149 n. 50, state that these doubts have been ‘laid to rest in PNDevon, pp.62-63′. So – to the library to discover what PNDevon yields, but not today because it’s raining.
One point to make before investigating this is that there are two separate points: Was arx Cynuit at Countisbury? Is the name Countisbury derived from Cynuit (or some form of it)? It is possible for Countisbury to have a different derivation, yet for arx Cynuit to have been on Wind Hill: ‘Cóndatisburh’ derived from a geographical feature, ‘Cynuit’ from an individual.
At any rate, Stevenson, an uneven scholar apparently, edited Asser’s Life of Alfred together with the Annals of St Neots, and in discussing the incident of 878 he considers the etymology of Cynuit and thought it ‘impossible phonetically’ for it to give Countisbury, as did I: (‘Countesbury, quasi Cynwitesbyrig could not have had the gen[itive] sing[ular] es‘). He toys with ‘an Old Welsh form’ Cunetio/Cynwydd and Cunetio ‘the older Celtic name for the River Kennet’ OE Cynete, as did I, but concludes that there is no River Kennet in Devon. A land that I have visited – but Stevenson, it appears, was wrong – and the answer lies in PNDevon, pp. 62-63.
So, one last thought: just how likely is it that my ‘Condatisburh’ is the real derivation? I asked Google if it had any ideas: