I think most of the overshoot problem comes from the thermistor bead's tenuous contact with the nozzle: it's held in place by a wrap of Kapton tape and the fiberglass/silicone tube, so the heat transfer isn't very good. The nozzle gets hotter faster than the thermistor and the temperature overshoots while the bead is catching up.
I epoxied the thermistor bead and a thermocouple to the nozzle, then adjusted the thermistor table to make the thermistor temperature match the thermocouple. After that, the nozzle reaches 180 C in one minute, the overshoot amounts to 10 C, and the temperature stabilizes in about two minutes.... with the default 255 PWM setting.
I've been describing that process on my blog:
There's more and better data to come, but so far it's working pretty well...