Thanks for the reply.
I totally agree about censuses and the notion that an adult may have an inaccurate idea of where they were born.
Also agreed about DNA... to a point. I confess I only half understand this stuff, but I tend to use it as a clue to which trees to look at, and to where a common ancestor might be found. I appreciate that especially as you get further back, the best that might reasonably be said - as in this case - is that these DNA matches probably had ancestry going back to the same area and the same group of families. In some cases they may actually have specific ancestors in common, but it's not safe to assume this is the case. Nevertheless, it still seems useful to point you in the right direction if enough DNA matches also point that way I think.
I've always started from modern times and worked backwards through the paper trail as rigorously as possible, ignoring other people's trees and relying solely on the evidence. Sooner or late though the evidence runs out or there's no way of choosing between multiple possible baptisms, marriages,etc. - again, just as in this case. At that point I think it's worth looking at DNA matches, with all the provisos already mentioned taken into account, just in case it's possible to identify an ancestor further down the line and then work backwards to the brick wall. At least that's my theory! I don't know how often it can work in practice - certainly in this case whilst I'm confident that Wedmore is the epicentre of the various Vowles lineages, including mine, there are just so many Vowles in the area, all with similar names, that it seems impossible to confidently say 'this is the one'.
I'm hoping someone can prove me wrong however and either say unequivocally 'this is the one' or suggest a new line of enquiry!