It's just that from a more complete point-of-view, there are many other areas than rotation curves that depend on dark matter and has been modelled to really require the specific properties of DM that the lambda-CDM model has, and it's just a bit worrisome every time a blog post talks exclusively about rotation curves. Sounds good that he writes about more of the issues elsewhere! The real picture might require a new blend of both DM and MOND of course..