This seems plausible to me. I'm not a physicist, but I think the case of dark matter can be compared to the case of the planet Vulcan.
In the early 20th century, in the heyday of discovering planets by observing existing planets and realizing that classical mechanics didn't quite predict their orbits correctly, it was observed that Mercury's orbit didn't fit the classical model well enough. Reasonably enough, it was presumed that there was another planet orbiting inside of Mercury's orbit. This planet was dubbed "Vulcan", and the fact that no one had ever observed Vulcan was explained away by the fact that it was clearly too close to the sun to be observed. This was all well and good until the early heyday of Einstein's theories, which made small but crucial changes to the mathematical models of classical mechanics. The orbit of Mercury was found to be perfectly consistent with Einstein's model.
The question of dark matter supposes a dichotomy: either our current understanding of physics is essentially accurate and there's a ton of fundamentally unobservable matter in the universe that explains any anomalous observations, or our model of physics isn't quite right yet. Intuitively, I think it's more likely our physics are off.
Either way our model of physics is off. Currently, the most reasonable explanation for dark matter is that it consists of some unknown, weakly-interacting particle. However, the standard model of physics predicts the existance of no such particle. Even this, the most mundane possibility, requires radical changes to our most fundamental physical theories.
That sounds even sillier. The most reasonable explanation is that there's some impossible matter around? At least the planet Vulcan was an ordinary planet that was just too close to the sun.
Again, I'm emphatically not a physicist. Stranger things have happened. It just smells funny to posit dark matter as a hypothesis.
The only thing that's not suspicious is the anomaly itself. Physics has always refined itself by looking at different scales. We were wrong about the quantum level the first time we looked at it, and we were wrong about even the planet Mercury, so it's to be expected that we're wrong about galaxy clusters.
"Impossible matter"? What's "impossible" about it? That we don't know what it is, and that it isn't predicted by our current models, doesn't make it "impossible." It means we're going to need to extend our models, which we've been doing on a regular basis for hundreds of years. Rob Knop sums this up nicely:
"In the face of evidence otherwise, many still insist that most of the Universe must be made up of baryonic stuff that interacts with other baryons and our familiar photons. Is this not just as much hubris as insisting that the Earth, where we live, must be the center about which all the other Solar System bodies orbit?" (http://scientopia.org/blogs/galacticinteractions/2011/08/14/...)
The dark matter hypothesis seems to have a great deal of explanatory power. The antigravity hypothesis, like other MOND theories, has a good fit for the specific phenomenon it was designed to tackle, but doesn't explain a lot of the other things that dark matter does. Ethan Siegel has a good post on "What Dark Matter's Alternatives Must Do" (http://scienceblogs.com/startswithabang/2011/08/what_dark_ma...) which points out some of the stuff that dark matter explains and rival theories don't (e.g. temperature fluctuations, hydrogen/helium ratio). He does presciently note that 'antimatter has negative mass' could explain more than traditional MOND theories but claims that this impacts other well-tested assumptions such as conservation of energy.
Of course, none of this is to say that the dark matter hypothesis won't eventually be disproved and go the way of Vulcan as you suggest. Ironically, though, it sounds to me like the MOND and antigravity theories are the Vulcans here: they solve only one specific problem, while dark matter fits with so much more.
That we don't know what it is, and that it isn't predicted by our current models, doesn't make it "impossible."
I was indulging in a bit of poetic license with that one.
I'll agree that we don't currently have a good theoretical alternative to the dark matter hypothesis. Much smarter and more qualified people than me have tried and failed to come up with one. But it still seems unsatisfying to me.
In the early 20th century, in the heyday of discovering planets by observing existing planets and realizing that classical mechanics didn't quite predict their orbits correctly, it was observed that Mercury's orbit didn't fit the classical model well enough. Reasonably enough, it was presumed that there was another planet orbiting inside of Mercury's orbit. This planet was dubbed "Vulcan", and the fact that no one had ever observed Vulcan was explained away by the fact that it was clearly too close to the sun to be observed. This was all well and good until the early heyday of Einstein's theories, which made small but crucial changes to the mathematical models of classical mechanics. The orbit of Mercury was found to be perfectly consistent with Einstein's model.
The question of dark matter supposes a dichotomy: either our current understanding of physics is essentially accurate and there's a ton of fundamentally unobservable matter in the universe that explains any anomalous observations, or our model of physics isn't quite right yet. Intuitively, I think it's more likely our physics are off.