Hi Ken,
What I meant when I was responding to Armin was not-so-much tongue-in-cheek (though I meant no offense) as it was trying to point out that the all-at-once paradigm doesn't make any sense to me when coupled with the concept of "updating." The former implies a singular while the latter implies a plural.
So here's the way I see it. What you could be saying is that we do an "all-at-once" analysis of the universe and then as we learn more information about it (both past and future) we continue to do so and thus "update" our information. In this way the underlying universe remains essentially "static" while our knowledge of the universe possesses an arrow of sorts (since we are gaining information about it). Now, I'm not particularly familiar with Matt's notion of contextuality and I'll have to do some digging into it, but from my own notion of contextuality there's a problem: the universe does *not* remain static. In other words, it is not always possible to deduce *past* states from future ones. It would seem that your argument does not address this issue.
Now, regarding the time-asymmetry issue, I disagree with Carroll, Penrose, et. al. Rather, I agree with Eddington who did not, necessarily, postulate new physics to explain this problem (he postulated new physics to address other issues). Eddington viewed the time-asymmetry question essentially entirely in terms of probabilities (actually, that's pretty much how he viewed everything). I really don't think it is that mysterious. The problem with the Carroll, Penrose, et. al. interpretation is that it assumes that the Standard Model is correct which in turn says that CPT-symmetry is inviolable. But there are notable problems with the Standard Model (it fails on a number of counts not the least of which is in relation to gravity). In addition, there is some evidence that CPT-symmetry is no inviolable. Recent studies of neutrinos suggest that it is possible that neutrinos and anti-neutrinos have slightly different masses. If so, this would be evidence of CPT-symmetry violation. All of this is to say that I think the cosmological case for the arrow of time is built upon some shaky ground (and I didn't even mention inflation). Heck, Sean Carroll himself addressed some of these issues at the last FQXi conference.
Now, to the c-frame issue. Obviously you are correct in saying the the c-frame itself is unattainable and, perhaps, unphysical in some regards (though clearly it is at least partly physical since light exists!). But the bigger point is that there *is* an objective ontology to relativistic frames. Otherwise, the Twin Paradox would really be a paradox. But it's not. The space-faring twin really is older --- they both agree on that when they meet again. Likewise, relativistic muons created in the upper atmosphere really do take longer to decay. And as two frames get closer and closer to c, they get more and more in agreement. That's the objectivity that I'm talking about that we all tend to ignore. If reality were completely subjective (relative), then the twins would each think the other was older.
This all goes back to that deep issue of time. While the laws of physics (mostly) seem to be time-symmetric, there's clearly a problem since we never see them run in reverse. If this is a purely cosmological problem, why is it that we never see, say, sub-atomic particles "un-decay?" Time is simply different. Relativity provides us with a convenient framework within which to treat time in a similar manner as space, but even *it* has a preferential arrow built into it --- allowing for complete time-symmetry in relativity produces logically absurd results. This is precisely because time is different in the metric --- it's sign is always opposite that of space. Incidentally, Lev Okun wrote an article about this very point a number of years ago but the cosmologists and many high-energy physicists still seem to cling to this notion that space and time are the same (and thus mass and energy also are the same), that the universe is really symmetric, that the Standard Model is correct, and that time-asymmetry is just an illusion perpetrated by cosmological expansion.
I hope that answers some of the points you have raised. If I missed responding to something, let me know.
Cheers,
Ian
P.S. It is thankfully much cooler today...