“The observations do contradict GR” and “the observations are at odds with GR under the assumption that the universe is made of ordinary matter” are two very different statements, and the latter does not imply the former.
You could also say that deviations in Uranus’s orbit contradict GR under the assumption that Neptune doesn’t exist. But Neptune does exist, so this isn’t really a statement about GR at all.
The difference is that Neptune is there, indeed, was discovered because it was exactly where Newtonian mechanics predicted it would be. The situation here is completely different because all of the plausible hypotheses about the nature of dark matter have been tested and falsified. The only thing left is "some weird stuff that is fundamentally unlike anything we have ever observed before". It is more analogous to the luminiferous aether than to Neptune. If that doesn't count as contradicting GR, at least potentially, then nothing contradicts anything because you can always resort to this kind of special pleading to explain any observation under any hypothesis.
I'm not suggesting it, those were literally the exact words that I wrote.
Of course, that claim turns entirely on the meaning of "plausible", so I'll amend my claim to simply say that all attempts to detect dark matter to date have failed.
I’m no expert in this area, but my understanding is that weakly-interacting particles aren’t “something we’ve never seen before”, and in fact the neutrino is an excellent example of such a particle. However the neutrino isn’t a candidate for the specific particle causing the observed effect because the distribution of observed effects on normal matter would look different, hence the search for more massive particles: a search that is very much at its early stages.
> weakly-interacting particles aren’t “something we’ve never seen before”, and in fact the neutrino is an excellent example of such a particle.
That's right. But the only reason we know about the neutrino is that it was predicted by theory.
> However the neutrino isn’t a candidate for the specific particle causing the observed effect because the distribution of observed effects on normal matter would look different, hence the search for more massive particles:
Correct again.
> a search that is very much at its early stages.
That depends on what you consider "early". There are only so many places in the standard model that wimps can be hiding, and the vast majority of them (if not all of them -- I haven't been keeping up with the latest developments) have been ruled out.
You could also say that deviations in Uranus’s orbit contradict GR under the assumption that Neptune doesn’t exist. But Neptune does exist, so this isn’t really a statement about GR at all.