I think that's been overstated. The fossil record is always incomplete, and not all evolutionary changes can be detected through the fossil record. That's why there's been so much interest in observing evolution in controlled environments that isn't limited by the obscuration of distant time.
And I don't see that Krauss was outmatched in this area (fwiw, Krauss is a theoretical physicist, which means he studies more than just cosmology). The fact that Craig presents answers to problems we currently don't have answers for is not him "beating" Krauss at his own game. Krauss is dead on when he says "we don't know." It's that simple. Craig would like to convince you that he knows something Krauss and other theoretical physicists don't on the subject of physics in cosmology. He doesn't. That's why Krauss and others like him are unimpressed.
It's true we may be talking about the limits of our mathematical models when we talk about infinity, but we currently have no other word for it. Read my replies to stuntpickle about infinity: "counting an infinite number of things" presumes that infinity obeys the same laws as finite things. It's possible (likely) that there can't be "an infinite number of (discrete) things," because suggesting discreteness already presumes finiteness. Krauss is right in that all physicists and mathematicians "use" infinity and it does seem to function and have its uses. As for what that means regarding real-world referents is debatable, but Craig simply saying that it can't exist isn't convincing, nor is his trying to demonstrate the absurdity of infinity using finite calculations.
Yes, but see the problems I outlined above.
See above again on the problem of "nothing". All we can really say is that whatever created the universe wasn't limited by spacetime, and must have brought both into existence. But we still don't know what that thing is. It doesn't really supply support for the Judeo-Christo-Islamic theory of "something from nothing," because we don't know if it was nothing. It's still possible (maybe even likely) that it was quantum gravity. Again, let's wait for some tests to provide some kind of solid conclusion before we accept another theory that gives us no predictive power.
To validate the position there would first have to be a way to falsify it, and there's not.
Let me add something I haven't brought up yet: In mathematical logic there's something called the Conjunction Fallacy. Simply stated, if one probably is contained inside another one, the one that's contained will always be more likely to occur. So, if you look at two dice-roll sequences:
1. 531542
2. 3425651
The first sequence will ALWAYS be more likely because it is contained inside the second probability (the 531542 sequence occurs in the second, out of order, with a 6 added in an extra roll).
What this has to do with the Kalam is this: Let's accept that the universe must have a cause that precedes spacetime. Let's suppose that this cause could be narrowed down to God or quantum energy. All things being equal, we should always state the latter is more likely. Why? Because we know quantum energy exists, whether God exists or not. So even if God does exist, quantum energy fits inside that hypothesis the same way the 1st roll in the dice example fits inside the 2nd.
This is often sometimes informally expressed as "Occam's Razor", or "the simplest answer is usually the best." Every extra element we add to a theory adds something that can go wrong with it. God "seems" like a simple answer until one gets past the linguistic simplicity to imagining rendering ourselves in, say, AI terminology, but adding abilities that even we don't possess. In comparison, utilizing things we already know exist (like quantum energy) is to be much preferred, if only for the reason we don't have to worry about the probability of its existing at all.