What's the evidence, if any, for some local concentration(s) of dark matter in some region(s) smaller than a galaxy? What's the evidence, if any, for some local concentration(s) of dark matter in some region(s) smaller than a galaxy?
Galactic-sized - or larger - gravitating halos seem to get all the attention.  I'm just wondering about the status of smaller halos/concentrations, if any.
 A: This is an active (hot?) topic of research, in fact I attended a workshop on the subject just last week. In brief, no one has found a dark matter (DM) halo yet that does not host a galaxy, though we would very very much like to!
The first reason it's so difficult to find a DM halo that does not have a galaxy is that a common working definition of a galaxy is "a collection of stars, gas, dust (possibly other things, e.g. super-massive black hole) that sits in a dark matter halo". This is in contrast to star clusters (globulars, open clusters) which are collections of stars thought to be devoid of significant concentrations of DM. Traditionally, dwarf galaxies and globular clusters were classified separately, and the DM content was thought to be one of the defining characteristics. Now the delimitation between the two classes seems to be blurring a bit, and it's not clear whether these ideas and definitions will hold up in the long run. So the tricky part is that to find a DM halo without a galaxy in it, you need to find one that is actually empty, if there are any stars (or, arguably, even a gas cloud) what you have is a galaxy.
Another complication is that any such halo must be very small, less than about $10^{9.5}\,{\rm M}_\odot$, give or take, because every more massive halo is expected to host a galaxy. Since the only way to detect an empty DM halo is by a gravitational effect (not quite true, see below about self-interacting DM), you need to see either a gravitational interaction, or a gravitational lensing event. Because the mass is low, the possible effects are rather weak. Some current leading ideas:


*

*A dark DM halo colliding with a stellar stream could leave a tell-tale gap in the stream. This is a very subtle effect, and is very difficult to both measure and model. And on top of that, it's not supposed to be especially common, and since stellar streams are only observable in sufficient detail for this measurement in our own Milky Way galaxy, there's a chance that none of the streams will currently have a gap and we'll just be out of luck.

*A dark halo merging with a dwarf galaxy could be pretty spectacular - it would look rather like an ordinary major merger of two dwarf galaxies, except only one of the merging dwarfs would be visible. The merger rate isn't expected to be terribly high, and one has to be careful not to get confused with other types of disturbances to the system. In this case it's also very hard to determine the mass of the dark halo, which is a quantity of considerable interest.

*In strong lensing systems, a gravitational lens at intermediate distance creates multiple images of a background object. For a "simple" lens like a massive elliptical galaxy, there is fairly good symmetry and the images can appear in a handful of known configurations. Based on their relative orientations, one can construct a very strong prediction for the relative flux (brightness) of the different images. Departures from the predicted "flux ratios" could be caused by perturbations to the lens by dark substructures. This is called the "flux ratio anomaly", and a handful of them have been observed. The tricky part is convincing ones self that the FRA is due to a dark halo and not something else (globular cluster or dwarf galaxy perturbing the lens, hard to check this because the lenses are so far away); or intervening dust cloud blocking some light from one of the images.

*One of the key interesting results from the workshop I was at was that the strong lens models are beginning to be sophisticated enough, and the observations detailed enough, that it might be possible to not only measure the presence and position of a perturbation to the lens, but also its distance. This is key, since a perturbation outside the lens system (i.e. out in some generally low density chunk of Universe) is unlikely to be a globular cluster, and if the mass is low enough to rule out a dwarf galaxy, this may be the best chance at a detection of a dark halo in the near future.


The implications for a detection are substantial. The detection of even a single low mass ($10^6\,{\rm M}_\odot$, potentially even up to $10^7\,{\rm M}_\odot$) in the right circumstances is enough to rule out all $7\,{\rm keV}$ sterile neutrino warm dark matter models, which are currently the state-of-the-art WDM model. A sufficiently convincing non-detection would be very difficult to reconcile with standard cold dark matter.
Some additional footnotes in the form of disjoint paragraphs:
It's worth briefly mentioning self-interacting dark matter models. If dark matter can self-annihilate, then we might expect to see $\gamma$-rays coming from the centers of halos. Some (contradictory!) detections have been claimed, but we'll have to wait and see. In some models, the strongest emission is expected to come from low-mass, dark halos. Gamma rays coming from "nowhere" would be a pretty tell-tale signal, if detected.
Another take on DM structures smaller than entire galactic halos is "dark disks". This is exactly what it sounds like, a disk of denser dark matter that lies roughly coplanar with a galactic stellar disk. People have argued that the Milky Way has one of these, and even seriously argued that it killed the dinosaurs. I'm (shameless plug) a collaborator on a paper that is soon to appear arguing that the MW is unlikely to have such a disk, and that even if there were a disk, it is of little consequence for experiments attempting to detect DM via nuclear recoil. I don't know about the dinosaur thing, though. Maybe I should read the book ;)
Finally, on a very prosaic note, all massive galaxy halos have dark matter substructure orbiting them in the form of little clumps, but all those detected contain a dwarf galaxy. The same goes for galaxy clusters, but the subhalos can host fairly massive galaxies (or more dwarfs).
