http://arxiv.org/abs/1503.07675 claims they tested some colliding galaxy clusters and found the gas got moved, but the mass got moved by a different amount than the gas. Their techniques for measuring this seem remarkably sensitive (if we believe them). This "proves" dark matter exists [claimed p-level 10^(-14) or so]. This contrasts with earlier "proofs" which involved the rotation curves of galaxies. Two proofs are more convincing than one, and alternate "new physics" explanations (different from assuming dark matter exists, that is) now may have more trouble surviving. They then go further and estimate the interaction cross-section of dark matter particles with other dark (or non-dark) matter particles. They claim it is too small for them to detect, less than 0.47 square-cm per gram. If this can be brought below 0.1 this may actually disprove the existence of DM (!?) they say, because they say other observations of DM suggest it must self-interact. So this whole line looks very promising, at least on its face.