Llano was 32nm, was it not? AMD mentioned with the Phenom II's (45nm) 1.55v was the most they'd recommend for vcore. I've seen some documentation for BD and PD that indicate the same thing actually. I wouldn't go over that, not unless you're ok with the fact that your CPU might not last as long as you want it to.
I've only heard of a few fried CPU's from electron migration, but each one of those that did was using too much voltage. I think about the fastest I've seen a chip die was a year or so (was an old windsor/brisbane chip on 1.65- 1.7v). I've also seen chips that have started to show signs of electron migration (no longer holding the OC with the same voltage, meaning, it required more and more voltage to keep the same speed). Most of those cases have ended with people upgrading the CPU before it technically died, but I think it still counts.
Killing a processor has nothing to do with temperature, and everything to voltage. Extreme overclocking is done with liquid nitrogen and helium (-250c) and extreme voltage (1.8-1.9v). Extreme overclockers also regularly kill chips. Clearly it is not temperatures that are destroying the CPU, but it is instead voltage.
And yea, I'd be suspect about your temp readings being inaccurate.
Good luck on the OC though