Very good question, and unfortunately I don't know the magic formula, obviously they are connected. There are lots of methods how we measure stability: Bode, Nyquist, Barkhausen, Routh-Hurwitz, Rollet, etc., we don't use these at the same time, I guess because distributed-element models and lumped-element models are far enough from each other in the term of stability, so we cannot use easy calculation between them. You couldn't implement measurement method which is good for non-RF and RF circuits. S parameter measurement is good for RF, but not good for non-RF, analog circuit. Routh-Hurwitz is better for a complex control system than for an analog circuit, and so on...