I like to be able to see a tiny sliver of my car. This way I always know how they're adjusted. Without being able to see that piece of the car, they might be adjusted too far out and I have no easy way to tell.
well, if you see a sliver of your car, you know how they're adjusted: wrong :)
seriously though, it's really easy to tell, just watch a car that is about to pass you transition across your mirrors into your peripheral vision. if there are any blind spots or periods where the car is highly visible in more than one visual source (mirror or peripheral vision) then you are seeing less than you are able to. i'm pretty sure there is no free lunch here -- if you are still able to see your car, then there is likely overlap between your rear view and your side mirror's view frustum, not to mention there is unused visual field in the side mirror anyway, so you're sacrificing visibility in your side mirror.
(you should still check for the presence of a car by looking behind you if possible, or cross checking your peripheral vision with your mirrors, but it's nice to know that you don't have any blind spots if you fail to do so.)
You keep repeating this theory that the 'new' way for adjusting mirrors solves all problems, and simply ignore that it introduces different blind spots that small vehicles (such as motorbikes or bicycles) can disappear into.
Given that the article is about cyclist safety, and how it's improved by training drivers to look behind them in their blind spot, I find the irony of you suggesting this 'new' method to be amazing.
