There are real reasons to take, say, supersymmetry or grand unification seriously other than their usefulness in solving the cosmological questions you’ve mentioned. As I alluded to before: despite its experimental success, the Standard Model has known theoretical inadequacies (doesn’t account for the particle hierarchy without fine tuning, doesn’t explain apparent coincidences in the running coupling constants, is inconsistent with the gravitational interaction, etc.) Thus, appealing to “minimal modifications of the Standard Model” isn’t necessarily the conservative choice, as it retains all those problems or makes them worse. For that reason, those theories, while speculative, are often not regarded as particularly extraordinary nowadays: after being around for decades, they remain the most conservative modifications to the Standard Model which resolve some of these problems. One may appeal to cosmological solutions within the “extraordinary” frameworks of SUSY or GUTs or whatnot not merely because they are radical or new or “sexy”, but because they kill two birds with one stone (solving a cosmological problem and fixing the Standard Model).

(One can also argue whether “theoretical inadequacies”, as opposed to disagreement with experimental evidence, is a valid scientific reason to discount proposals which retain those inadequacies.)

A defense of not-so-extraordinary extraordinariness out of the way, I don’t want to lose sight of the fact that, as I have mentioned, there are a number of long-standing and plausible solutions to these cosmological problems along the “minimal modification” lines you listed. Nor is it the case that “radical” solutions are preferred for the sake of merely being radical: the merely “extraordinary” proposals such as SUSY or GUTs, which have been around for a long time, still have far more mainstream acceptance than the newest and more radical proposals such as large extra dimensions, braneworlds, etc.

I’m looking forward to your discussion of media impressions. I suspect their validity varies between sources (e.g., Scientific American vs. New Scientist) and from field to field.

For that matter, I don’t know any popular media which actually cites physicists as preferring the most exotic scenarios; at best, there is simply more coverage of such topics. Pay attention to how often the quotes from physicists other than the inventors of the new theories say things like, “It’s an interesting idea but it’s far too new to properly evaluate” as opposed to “It’s one of the best and most plausible scenarios out there”.

Instead of shifting the burden to a survey which disproves “the impression [Robin Hansen] gets from the popular media”, I wonder how you justify holding that position in the first place.

String theory in particular has yet to take over high energy phenomenology; explanations for particle physics and cosmological conundrums, when they go beyond the Standard Model, are still more likely to be framed within the context of supersymmetry or grand unified theories. (Of course, both are part of string theory too, as is arguably the Standard Model itself, but they are not as radical as extra dimensions or braneworlds.)

Regarding your specific points, my short answers are:

1. There are so many inflaton candidates that none are really “preferred”, although candidates which solve problems with the Standard Model in addition to solving inflation are preferred to those which only solve inflation.

2. Electroweak CP violation exists but is difficult to use to explain matter-antimatter asymmetry, so there is a legitimate reason to go beyond the Standard Model in that case.

3. Axions are still important dark matter candidates.

4. I don’t buy “dark energy is magnetic fields” at all.

Elaborating on point 1: Scalar inflatons have been the preferred mechanism for inflation for decades, although string theory gives you some exotic new possibilities for what those scalars could be. Actually, most inflation theories simply require an inflaton potential with a particular shape; whether you get that potential from an appropriately chosen new particle added to the Standard Model or from some string theory model is up to you.

I will note, though, that there is certainly a push to propose new inflaton candidates that are “natural” within some theoretical framework, as opposed to a random particle with ad hoc properties glued onto the Standard Model. So those scalars which appear naturally in supersymmetric theories, grand unified theories, string theories, etc. are popular.

(It is harder to get an inflaton to appear as a “natural” addition to the Standard Model, because the Standard Model doesn’t really explain how you can get the high energy field values necessary for a sufficiently flat inflaton potential. You can just add such a particle in by hand, but it doesn’t really “fit” with the rest of the theory, which operates at a very different energy scale.)

Note also that many of those theoretical frameworks (SUSY, GUTs, etc.) were themselves proposed in order to solve some long-standing issue with the Standard Model, such as the hierarchy problem, or explaining the extrapolated convergence of running coupling constants — in other words, there are legitimate reasons for going beyond the Standard Model.

That being said, I will return to my earlier point that there is an enormous collection of inflaton candidates coming from practically every source you can imagine, including simple additions to the Standard Model, and that people have mostly retreated from promoting specific scenarios to trying to phenomenologically characterize the parameter space of inflation theories. There are simply too many workable proposals out there.

On point 2: CP-violating interactions are certainly part of the explanation of matter-antimatter asymmetry, but the question is what mechanism is at work. Ordinary electroweak interactions don’t provide a large enough asymmetry. You can extend the Standard Model in various minimal ways to get the required asymmetry, but there are not too many parameter values which work there either. See this review.

On point 3: Axions are a very nice solution to dark matter, and there are good independent reasons to believe they exist. But it’s tricky to get them to have the right mass to serve all the necessary roles of dark matter (see here). Dark matter needs to be compatible with galactic rotation curves, early universe structure formation, and CMBR anisotropies, among others. It’s hard to satisfy all of those constraints at once. It’s hard (but not impossible) to get axions to satisfy them all, but it’s also possible that axions are just one kind of dark matter and that others exist too.

On 4: Dark energy as magnetic fields doesn’t seem at all plausible to me, but I will defer to experts in the field (none of whom have weighed in on this subject). It doesn’t help that it has been covered by New Scientist — judging from their coverage in fields in which I am qualified, NS has descended to near-crackpot status in the last few years, and I automatically downweight almost everything they publish nowadays. (Witness their recent sad promotion of research which purports to violate the conservation of momentum, incorrectly claiming that this is theoretically permitted by special relativity.)

But then the history of science shows that, just occaisionally, very radical theoretical changes are needed to to explain the natural evidence. Two outstanding examples are the change from the Ptolemaic Earth centred universe to the Copernican theorry of a spinning Earth as one amongst the other planets orbiting the sun, and Planck’s theory of quanta, which at least appeared to contradict the existing wave theory of radiant energy and led to the development of the quantum theory of the standard model.

So that one could think, perhaps, that the general problem of developing any experimetally testable theory of quantum gravity like superstring theory is that this approach to finding a successful ‘theory of everything’ is not radical enough.

Which shows that, no matter how often science is wrong, on any randomly-chosen question it’s very likely right.