Greebo, it seems you completely missed the point I was trying to make. Let me try and explain a little further.
Yes but the same applies to temps as well so you are wrong on that front. You will get variation between chips with different temps for same VID but there is a definite link between voltage and temps.
Agreed, there is a definite link between voltage and temps, but not across VIDs. i.e 1.4v on a 1.2VID chip is generally going to be hotter than 1.4v on a 1.35VID chip (+0.2 and +0.05v respectively).
Take your example. Even if the 1.20VID chip runs hot due to poor ihs at 1.2v it will still run cooler at 1.4v compared with another running at 1.55v
Not at all. At 1.2v, on a stock Intel heatsink, i think this Q6600 would throttle itself at over 70C and i wouldn't be surprised to see 80C. I'm fairly sure you can find some Q6600's that will take 1.55v on a stock cooler. Temperature is very chip dependant, and not linked to VID at all. A lot of people will blame bad mounts, but the truth is that unless you're doing it fundamentally wrong, you'll only see a 2-3C range in relative good/bad mounts.
My 1.25v chip runs at 40 degrees under load at 1.2v, 50 degrees at 1.3v and hits mid 60's at 1.55v.
Further backs up my claim.
Lower voltage is better for overclocking potential and some people run out of headroom due to poor cooling and find their temps limit their overclock. But that is only cause they are running at such high voltages so hence your second bit is correct - been able to run at 3.4Ghz @ 1.4v is better than 1.55v, not because of the voltage but because of the heat.
Lower voltage, combined with lower temps are good for overclocking. If you have to sacrifice one, you sacrifice the insistence on low voltage and keep low temps.
I would be happier with a 1.35VID chip at 1.55v running 3.6Ghz than a 1.20VID at 1.45v running 3.6Ghz given both temps the same below 60C.
In saying all of that, you still get good and bad overclocking chips of the same low VID, it's just that your odds of getting a better overclock increases with a lower VID.
Agreed. The main reason is that people feel happier with the volts reported. If you look at the voltage increase (i.e +0.2v or w/e) then you're likely to see the same trends (taking averages of good/bad chips).
Take a look on the quad core overclocking thread. The lower VID's are in general nearer the top of the charts but some aren't and conversely so high VID chips are up there.
True, exactly as I've explained. If 1.6v is considered max on air, then 1.6v is +0.4v on a 1.20VID or only +0.25v on a 1.35VID.
As you'll see, I'm a firm believer in voltage increase being more important than actual voltage.
Basically 70C... 
...
bleeding huge cooler and still no joy