The vertical axis is cross-section of interaction, the horizontal axis is mass in GeV/c^2.
LUX pretty definitively rules out the possibility, hinted at by several dark matter experiments, of a dark matter particle in the 5 – 20 GeV/c² mass range. While XENON100 seemed to contradict this possibility already, it didn’t do so by a huge factor, so there were questions raised as to whether their result was convincing. But the sort of ~10 GeV/c² dark matter that people were talking about is ruled out by LUX by such a large factor that finding ways around their result seems nigh impossible. . . . by 2015 their results should improve by another factor of 5 or so.
From here (some parenthetical matter omitted). Lubos concurs with Matt Strassler's analysis above. He states:
The exclusion and the agreement with the background expectations [in the LUX experiment] is spectacular (no phantom events appear in their data at all) and I have no doubts that all the "Yes signals" [in other direct dark matter detection experiments] are due to some misunderstood background. This result hasn't excluded all meaningful models of dark matter yet but upgrades of this experiment are able to get to the region where the cross sections start to become unnaturally small.
Physics blogger Jester is also convinced.
Direct dark matter detection experiments already ruled out heavier dark matter particles up to about 1000 GeV/c^2.
This result confirms indirect astronomy inferences that rule out WIMP dark matter as heavy as 5 GeV or more, i.e. Cold Dark Matter, while not ruling out profoundly lighter "Warm Dark Matter" particles of ca. 2 keV (each of which would be about 2,500,000 times lighter).
The exclusion ranges rule out cross-sections of interactions comparable to those of neutrinos, (also here) which would be an a priori expectation for a weakly interacting dark matter particle.
Of course, the theoretical extreme of "collisionless" dark matter would be impossible to detect, by definition, in an experiment like LUX. Even an almost collisionless dark matter particle like a "sterile neutrino" which does not interact via anything other than gravity and Fermi contact interactions, would likewise have such a tiny cross section of interaction that this kind of experiment could probably not detect these particles.
This exclusion also highly constrains the mass of any lightest or next to lightest superpartner in a supersymmetry theory that could be a dark matter candidate, effectively closing a significant share of SUSY parameter space. The LHC rules out light superpartners up to the 100s or 1000s of GeVs. LUX rules out SUSY WIMPs of 5 GeV to about 1000 GeV. SUSY models with, e.g., a stable (or nearly stable) very light gravitino, in which all other superpartners are very heavy, are increasingly hard to construct, particularly if they are at all "natural."
We have long known that "hot dark matter" with particles of mass on the order of 1 eV or less, are excluded by astronomy data. Thus, dark matter has to be heavier than the known neutrinos. Taken together with this direct detection exclusion, the range of permissible masses for dark matter particles, if they exist, has to be of the same order of magnitude as already known fundamental fermions (other than the top quark) and the known hadrons - probably at the very low end of this range.