I have, at last, readed the following article of Leonard Suskind on the antropic landscape:
Before briefly discusing it I will mention that I have also beeing reading the chpater on cosmology of dinés book "supersymmetry and string theory" and some stuff on blogs and forums about cosmology. I must say first that a few years ago (maybe a decade) the view of cosmology among theoretical physics (or at least the ones I speaked too) was diferent that what actually is. The viewpoint was that there were physic firmely based on carefull and exaustives experiments made in earth, which were the core of physics and that as an interesting toy you could try to aply that laws to the universe as a whole and to see how well it fitted. The consensous was that the important aspect was that the universe could be coarse grained described by a, relatively, very simple models. If there was some particular deviation it was a good thing to try to improve the models to fit it, but it was not a preferent problem.
A good reason for not taking too seriously cosmology is that the universe as a whole is a very bad experiment. There are too many possible variables (some of them probably ignored) playing in complicated manners. In an earth based experiment you can modify the settings so that only a few parameters are relevant for the quantity you try to messure, but you can´t do something similar with the whole universe.
But from a time to now the viewpoint has drastically changed. The string theory comunity have the viewpoint that it is very unlikely to see quantum gravity effects in colliders (an exception could arise if brane universe scenaries are correct) and, in general, another stringy effects are also hard to see. On the countrary the early universe, with it´s high energy, would be the place to search for this efffects. I must´say that I have become surprised when I have seen some of the effort made in this line of research. To say one I never even had thoguht that someone could have cared about the possibility that the decayof a gravitino, the supertpartner of the graviton, could be dominant at the same time that baryogénesis and that it could rsult in breaking of light elements. Or that a similar situtation could exist fo rmoduli fields of supersymmetry or string theory. There are workarounds for this posibilities, of course, but, what the hell? Who cares?. I mean, the interest is to find signs of supersymmetry now, not in some scenaries that could have existed, but that are excluded by observation.
But, ok, If some people wants to work in that things is fine for me. Even better is that people could be using string theory to calculate potentials for inflaton fields when they previously introduced that potentials by hand to try to fit observaations. The LQG comunity are using extensively the canonical (hamiltonian) aproach to do a lot of toy models. Irrespectively of how good or bad you thnk about LQG their effort is coherent in their internal logic.
In general I don´t dislike when people are triying to use fundamental physics (or should be physics,i.e. non still proved theories like strings and LQG) to see how they can help to fit cosmological observations. My problem is when the flow is in the reverse direction, when people try to modify physics (or should be physic) froms cosmological observations. A previous aclaration is necessary. If some theory predicts an aboundancy of phenomena such as monopolesor cosmic strings and hthat boundance is not observed one, certainlly, need tobe cauptios with that particular model. That is, one could be aware that the model can´t be the final answer to unification physics (althought still could be that some factor is the cause of the nonobserved aboundances and the model still be valid). But one would not rule out a model, with physical consecuences of it observed in earth, because of cosmological situations. Nor one would leave the traditional way to do physic because or cosmological observations.
Well, as everybody who knows the actual state of string theory knows this is not the case. The observation of an small positive cosmological constant has hanged the way to do physic (or at least string theory). The problem with that cosmologial constant is that string theory, as well as any quantum field theory for what matters, has a problem to explain it. If one calculates for an ordinary QFT the vacuun energy one finds that it is aan few orders of magnitude higher than the observed cosmological constant. If one has a supersymmetric theory, with unbroken supersymmetry, bosonic and fermionic modes canell out and one can get a 0 cosmological constant. The "little" problem with it is that suposedly supersymmetry is broken.
Well, the problem has existed from a long time, would it be with 0 or with a tiny cosmological constant. And people has lived with the problem and has been happy doing physics. But now it results that string theorists had a good confidence that afther all string theory could fit well with a 0 cosmological constant. And when the have found that it could be non zero (the experimental status of this is good, with at least 3 partially unrelated phenomenae indicating it, but still there are possibilities that afther all it is 0) the have become nervous and have modified the way to thnk about string theory, and it´s vacuums, to accommodate the reslt. In fact they have beein inspired by a model made by Weinberg a few decades ago. A model who was nothing but a curiosity who none cared too much because the had more interesting things to study.
But now you must leave with the idea that the universe must be in a sperposition fo supersymmetric vacuums, with 0 CC, and nonsupersymmetric ones, with a big CC, so the promedium is the observed one. Even you must face the possibility that this low value is hold only in a, relatively small, part of the universe separated by horizonts from anothers. And the criteria (while someone finds something better) to make physics in this superposition of vacua is to use anthropic principles.
Well, nowadays even renuent people such like Lubos Motl, seems to accept this status. My viewpoint is that there are a lot of things to do in string theory (or out of it) which is mainly unafectd by this change of paradigma. So, beyond being to the expectative that somone finds a better solution to the CC problem, I wouldn´t care too much about that landscape and would worry about other things, at least if that is possible (and I guess it will be).