I've now had the advantage of looking at all marriages 1857 to 1800 (+ a few years later in several parishes) - the parish clerk was a witness in I estimate 60-70% of marriages - the other witness probably had some family connection especially if they signed with a mark.
Minors were generally indicated but not that common, widows I estimate were not always indicated, widowers I suspect even less so. Braddan seems very deficent in all 3 cases in the 1800to1810 period I've been working on.