One of the odd characteristics of dark matter, aside from being dark, is that it forms halos around galaxies, but within those halos the dark matter is distributed evenly; it doesn't collapse into clumpy matter. Stars and planets, on the other hand, clump and cluster. If there were sufficient numbers of rogue planets to account for dark matter, there would have to be thousands or millions of rogue planets for every star that we can see. At that density they couldn't stay hidden -- over cosmic timescales they would be attracting each other, forming clumps that would partially occlude stars, and turning into stellar nurseries with detectable frequency.

As someone else alluded, dust absorbs and re-emits light in a way that can be detected by analyzing the spectral lines. So gas clouds and dust clouds should still be detectable, if there are many molecules between us and a light source behind them.

So that could be a dense cloud, or a very, very large but diffuse cloud of gas or dust, and there are a couple of galaxies that rotate too quickly for the number of stars we can see, and don't have enough visible dust to explain why they don't fly apart.

We're talking about needing to find something like 20 times as much stuff in order for the universe to make sense with the physics we have. And while I expect finding more stuff will improve the signal-to-noise ratio, at some point we should be looking for new physics, because new physics is where we find ways to do things the old physics says are impossible.