The contribution to the lensing effect is very small, because the effect of gravitationly lensing is very small, and the possible error from multiple sources quite large. It does appear from lensing that galaxies have a significant nonradiating mass, but that’s a result of correlations from observations of many galaxies, not a single identifiable result. And it doesn’t tell us anything about the nature of the dark matter, i.e. whether it is normal matter that is clumped up in baryonic nonradiating matter (MACHOs), or some kind of supermassive nonrelativistic nonbaryonic matter (WIMPs), or some kind of non-Standard Model-described virtual particles, or what.
To address the question asked by the o.p. regarding the motion of galaxies, what Hubble and Humason observed (which was later codified into Hubble’s Law) was that distant galaxies are redshifted to a degree in proportion to their distance, indicating that things further away are moving faster. Hubble et al were looking at Cepheid variable stars (supergiant yellow Population I stars) at the extreme range of observability, far beyond the local structures.
Not only are individual local galaxies moving toward each other, but all galaxies in the Local Group, and indeed, those in the Local Supercluster (along with many other superclusters in the area) are all streaming at high velocities toward a structure known as The Great Attractor, a supermassive concentration of mass in the region of the Centaurus Supercluster. Since it is occluded by being almost directly opposite of the galactic core, we can’t observe it directly but we can observe the relative motion of nearby structures against the stochastic background. In addition to that, there are even larger structures, galactic filaments (so-called “Great Walls”) which are condensed around unseen mass structures (hypothesized to be dark matter, cosmic strings, higher order spacial discontinuities, or who knows what). These large scale structures are difficult to tease out, and it’s only been in the last couple of decades that we’ve had confirmation of anisentropic flow and larger structure in local superclusters.
Why is this possible? The predominant theory regarding the Hubble Flow (universal redshift) is that it is due to the expansion of spacetime itself; the space inbetween objects is expanding, and the rate of expansion is slow between local objects, but is compounded with distance such that far away objects are moving away much faster, and the gravitational energy of nearby objects keeps them close together, overwhelming this dispersal. We can’t observe this expansion directly because we can’t look at this “spacetime” plenum; we can only see the effect of it upon massive objects and make inferences from the tenants of General Relativity. Besides giving us a reasonable explanation for the the Hubble Flow it also resolves Olbers’ Paradox, which is a good thing because without it we’d never have a good night’s rest. This also gives us a satisfactory rationale for the 2.7K cosmic microwave background, it being the result of expansion suddenly allowing the universe to become transparent to photons which have proceded to fly around with abandon ever sense.
Stranger