The notional figure cited from a study quoted earlier in this thread, which IIRC was offered up by Hentor, was 0.0004% fraud.
In a state the size of Wisconsin, with about 5.7 million people, I’d expect to see about 23 fraudulent votes. In your range, then, over 49 would surprise me. I have no idea how they might break out along party lines.
I’d say if there’s a small town, where everyone legitimately knows everyone else, then anyone you see at the polls that you don’t recognize, is, almost by definition, at least worthy of a second look, no matter their color.
I have no idea what size of town Wright was talking about and no way to judge how valid his perception of knowing everyone in town is. There certainly exist towns in which that perception would be valid.