I think the stock PID values were arrived at via testing on a big green egg, which makes sense since that is what Bryan uses. The PID isn't a magic bullet, it's the 'next step' in precision control of a system. It's a 'learning' system that will reduce variation - but it will take longer to control a system with large variation vs a system with small variation.
With my big steel keg, which is very similar to a BGE (can use same grate, diffuser, etc) I've found that the PID is nowhere near as critical as controlling the airflow - the volume allowed in by the fan (when off) is too much to run at 225*. You might have too much fan (or too much secondary air from leakage) - you heat it up quickly, overshoot, then it has to cool back down, undershoot & repeat. If you fan is off, then on, then off, etc - PID isn't going to help you, you need a more stable system.
You need a system where almost all the air needed is delivered w/o the fan. Then the fan can slowly add air to get up to the setpoint, then keep it there. overshoot is bad....
