This argument seems to pop up on HN every time dark matter is discussed. Unfortunately, it stems from a fundamental misunderstanding of how modern cosmology works.

The exact spatial DM distribution is not a parameter of the cosmological standard model. Instead, you assume an initial condition (pretty much a smooth distribution with only quantum fluctuations, which are parameterized by one or two parameters), apply the laws of physics to evolve this state some 14 billion years and compare the statistics of the theoretical and observed distribution.

In fact, DM suffers from the dwarf galaxy problem. Our theory predicts more than we observe. It's a bit of a challenge for DM, and considering how fuzzy our knowledge of galaxy formation is and how much other evidence for DM there is, it's not falsifying it yet. But if this is cemented by future observations it might very well be a blow to DM.

Besides, DM was hypothesized after measuring the famous rotation curves of galaxies. DM could have been easily falsified if we made gravitational lensing measurements afterwards and didn't see any DM. But we saw DM. Same thing with baryionic accoustic oscillations in the cosmic microwave background angular power spectrum. We looked there after we already thought that DM should be there, and the power spectrum looks just like if there is DM. It could have easily falsified DM, but didn't. The list goes on. DM could have been falsified may times, but passed almost all tests so far. Well, except for the ones in the lab, unfortunately.

It's not as if dark matter doesn't follow any rules at all - hypothetically, it obeys gravity like any other matter. So you can't have arbitrary amounts in arbitrary positions - you have to explain what attracted the dark matter to that position.