I agree wholeheartedly with everything said above, and would add: the importance of approaching everything about diabetes empirically. Bernstein is relentlessly empirical: not, "this is what I know," but rather "this is what I've measured."
That mind set must inform everything we do concerning management and control. As we know, everyone's diabetes is individual and behaves differently. The only what to know what's right for you is to try things, measure and record the results, and analyze and compare to find out what works for you.