I've sometimes wondered about this. I've heard it argued that Wessex was possibly originally something of a mixed kingdom, and given its legendary founders were Cerdic and Cynric, may have had a significant level of British influence in its upper "aristocracy", and on place-name evidence that at least a level of Brythonic-speaking population remained.
Since Cerdic and Cynric are very possibly legendary I've never put too much importance on this (though it does interest me why in Alfred's time the Wessex ruling class were happy to explicitly trace their line to natively-named rulers), but the other day I properly realised that Caedwalla also has a British name. While I've known this before I never really noticed it given the earlier Cerdic and Cynric - but as I understand it, Caedwalla is a properly attested historical figure.
Is there a general consensus concerning this? That perhaps the Wessex ruling line did descend from British forebears? Or that the Wessex ruling line deliberately took some native names to ease the integration of a Brythonic-speaking populace? Or something entirely different?
Alligning with a British identity was not necessarily unknown for the early English. Kent is an interesting example: the Kentish kings who attest legal codices in the 6th and 7th centuries use the title Cantwara Cyningas, which identifies them with the Cantiaci, the pre-Roman British people who inhabited the region, and who likely reasserted that identity in the fifth century. Archaeological evidence of the arrival of the English in the 5th Century indicates a pattern of settlement often alongside or overlapping British communities, so it's not necessarily unlikely that Germanic leaders would try to insinuate themselves into their new communties. If, as according to the Anglo-Saxon Chronicle, Cerdic was indeed an ealdorman who then established his own kingdom, it could well be the case that he was already in part British, as intermarriage with pre-existing elites is a common method for new arrivals to gain legitimacy, seen extensively in England in the wake of 1066.
Cædwalla is an interesting example, as his parenthood is only attested in part. His father is purportedly Coenberht, one of the Gewisse, but his mother is unattested, and may well have been British. What is important to remember is that these records are subject to a not inconsiderable amount of gentle manipulation in order to create a semblance of legitimacy linking Alfred and his immediate ancestors with the very foundation of Wessex itself. In fact, the line back beyond Alfred's grandfather, Ecgberht, who seized the throne, possibly with external aid, in 802 is extremely muddy. Given the West Saxon propensity for internecine conflict recorded prior to this, any connection between Ecgberht and Cerdic may indeed be entirely fabricated, or at least manipulated, in order to legitimise what may well have been in essence a coup led by a distant cousin or simply a leading ealdorman.