Somewhere in my mind I have the feeling Canon might not be doing the 0.18 micron process at all, but instead going directly to the next one (0.12, 0.09 or 0.065 maybe?), and that is why they have such problems in getting it to roll.
It makes sense doing something like that when you have your own factories and do not want to upgrade them too often. Going to 0.18, just to do it all over again in 2 years? I don't think so. But, leapfrogging the competition and take them by surprise, and in the same time prolonging the active time of the process platform? Yes, I think I'll vote for that.