I've used the Cumpiano method from the start. A couple of years ago I did some pull out testing on trial tenons and posted the results here. Can't find the thread in the archives but basically the tenon failed in tension along the grain lines at a moderate torque without reinforcement. (Pictured below)

With inlayed maple or carbon fiber strips it I couldn't get the tenon to fail. They worked as a tension band and formed a very strong construct. I now inlay Maple strips in the tenon, flood the whole works with thin CA and feel reasonably safe. I'm sure other methods work equally well. I think the Cumpiano method has some advantage with higher neck angles as in archtops etc.
