MOND paper says Bullet Cluster poses problems for Lambda-CDM

[hadsed]

I was reading a recent paper on arXiv.org on a novel simulation of MOND, and surprisingly they addressed the Bullet Cluster, which I guess is one of the more serious problems for MOND. What was even more surprising was that they said this posed a problem for Lamda-CDM models as well. Here is the quote:

Perhaps the most prominent example of a cluster with a serious residual discrepancy in MOND is the bullet cluster [29]. In this system, the gravitational lensing of background galaxies indicates that the mass is offset from the X-ray plasma. This is the same residual mass discrepancy that is seen in all rich clusters. While the bullet cluster is frequently cited as evidence against MOND, it is also problematic for ΛCDM. The sub-clusters that compose the bullet cluster collided at a remarkably high velocity (∼ 4700 km s−1). This is exceedingly unlikely in ΛCDM, occurring with a probability of only a few parts in a billion [30]. In contrast, such high collision velocities are natural to MOND [31]. Taken at face value, the bullet cluster would seem to simultaneously support and falsify both theories with equal vigo

Page 3, third to last paragraph in this paper: http://arxiv.org/PS_cache/arxiv/pdf/1102/1102.3913v1.pdf

Can someone explain what exactly the problem is? I was under the assumption that the Bullet Cluster was explained pretty well by dark matter.

 

[matt.o]

That 4700 km/s is the shock velocity and is not the velocity of the subcluster. The shock velocity measures the relative velocity of the intracluster medium involved in the collision -- the "ambient" medium ahead of the shock also has some velocity component (around 1000 km/s towards the shock). Springel et al. http://arxiv.org/abs/astro-ph/0703232v2" showed that the subcluster velocity is more likely to be ~2600km/s.

Also, I think they have not understood what was done in the Jee & Komatsu paper they cite, which also has problems.

 

[zhermes]

The General Argument:
The dynamics of the bullet cluster clearly show a discrepancy between the majority of the mass (interpreted as dark matter) and the majority of the observable material (gas and light). MOND (to my knowledge) has no way of interpreting or explaining this.

This quotation:
Is discussing something completely different than what's usually discussed. This argument is suggesting that the collision itself shouldn't have taken place in Lambda-CDM, but is natural in MOND (I don't see why that would be the case).

1) The fundamental issue remains, which I don't think MOND can explain (or even address---as exemplified by the author's deferral of the issue).2) I'm very dubious of their statistics here... presumably they're using some velocity distribution of clusters and finding the probability that these will collide at said velocity. This wouldn't take into account the (entirely unknown) evolutionary history of the clusters or that region of the universe.
*** EDIT: See @Matt.o above for a more educated analysis of this point ***