1.) "Life as we know it" is an extremely important phrase. Ask a biologist what life is, and she will probably say something about cells and DNA. It is possible for other self-replicating combinations of molecules to exist, but not on Earth, because the Cell/DNA model has become so effective at wiping the competition out. It's actually really hard to come to a solid definition of what "life" is, but if you make it sufficiently broad, I suspect you could find life even in some universes where the fundamental constants differ significantly from ours.
2.) We have no idea what determines the fundamental constants. You cannot talk about the "odds" of having a universe like ours until you know the probability distributions for each of the fundamental constants. Look at the H2O on Earth, and you will find that a significant portion of it occurs at 0C and 100C. Of course that is not a coincidence - water often exists at those temperatures because energy is spent on converting state, not on altering temperature. It's an equilibrium point. So likewise, it's POSSIBLE that the fundamental constants (of multiple universes, if they exist) interact in ways that cause them to arrive at equilibrium points. One really fascinating theory promoting this idea (and any theory is as good as the next, since we have very little empirical evidence available to us) is Lee Smolin's argument that the singularities within blackholes are themselves universe phases. Thus, a kind of "natural selection" of black holes occurs, which leads to more and more universes with fundamental constants ideal for producing black holes (and thus close to the constants necessary for life).
3.) The anthropic principle is just way too convenient, and I fear it de-motivates scientists from looking deeper into issues that really need explaining. Roger Penrose argues that there is a surprising lack of parsimony on behalf of whatever forces (spiritual or natural) brought about our universe. While our universe appears to have inflated from an initial region with a phase-space-volume of 10^(10^123), it would have been more than sufficient to create a life-supporting universe (several! in fact) from a region with only a phase-space-volume of 10^(10^117). Additionally, Penrose points out that the odds of creating an exact duplicate of our solar system from the random coming together of gas and radiation (no Darwinian evolution necessary!) is only 1 in 10^(10^60). (Penrose calls it "utter chicken feed" compared to the odds of the Big Bang producing ANY universe of our size). So even when you invoke the anthropic principle, the question still remains - why is our (life-supporting) universe so large and so old, when we could have done with so much less?