Sometime I want to look into this in a bit more detail:
Dynamical simulations of the universe predict a "cosmic web" of dark matter filaments (carrying baryonic matter as well) draining into nodes (clusters) as the universe expands. Recent research (e.g. http://arxiv.org/abs/astro-ph/0509262) shows that we can sort of detect these filaments by looking at the dynamics of distant clouds. This is pretty much the Ly-alpha forest.
In our own group of galaxies, we might be able to see this mass draining out of the filaments and into our mini-node (are we really a node, or more of a knot along a filament?). Maybe the high-velocity extragalactic clouds are really filamentary material? (http://www.atnf.csiro.au/pasa/17_1/putman/paper/node3.html and then http://www.aas.org/publications/baas/v28n4/aas189/abs/S061002.html -- notice the lack of refereed papers on this so far.)
There are, however, papers (ADS) showing that we are starting to have a decent handle on dark matter in the local group, so resolving the filament structure might not be that far off.
And once you start looking at extragalactic gas in detail, you start noticing some really odd stuff. (http://arxiv.org/abs/astro-ph/0409359)