Do keep in mind that while not allowing private communication avoids having mass claims by PM, it doesn't do anything about public mass claims. Having everyone claim in thread, then focusing the lynch on anyone who refuses or doubles up can still potentially break a game.
A vig that is at least as good at finding wolves as the lynch voting is makes the game faster, but shifts the ratio of kills in the favor of ones that can actually kill wolves. Whether the person who gets the role is good enough at scum-hunting for that to work out is still a toss up, of course. So I would probably be more of the philosophy that vig kills should be used sparingly.
I would say though that judicious use of vig power can be helpful even if the vig isn't confident in their scum-hunting. If the next day's lynch is a foregone conclusion (if someone got caught in a fake-claim, as a clear example), then vigging the inevitable victim skips a day that wouldn't provide much information anyway and so essentially denies the wolves an NK. Somewhat less clear-cut, if there is a complex day with competing wagons that go back and forth before finally lynching a townie, and so potentially has a lot of information to be gained from who voted where and when, vigging the other wagon gives a quicker answer to whether the competing wagons were town-town or town-wolf.
So, I think the information side of things could be just as important to consider as simply the pure likelihood of the target being a wolf. If someone's death gives town useful information, without losing too much from not having the voting pattern, then it can be worth it to essentially skip an NK.