And, if our theories are right, we're at 4.9% regular matter and 26.8% dark matter, so dark matter is five times as much as regular matter, so that's a lot of aliens...
Dark matter isn't non-luminous matter. It's matter that only interacts gravitationally, but not electromagnetically. This means it doesn't undergo collisions and can't shed angular momentum. It forms a diffuse, largely uniform cloud throughout galaxies. The result is that galaxies are more dense further from the galactic core than we would predict from luminous matter alone.
Dark matter corresponds to option 3 - and there are observations that conform some models of dark matter distribution that match quite well between different galaxies. There are other theories as well, such as MOND (modified Newtonian gravity) that explore option 2 (GR is wrong).
Still, whatever the theory, it's clear that what is not happening is "aliens someone consuming all of the EM radiation from some stars". With anything resembling currently known physics, it's impossible to "consume" EM radiation in this way. Electric charge is always conserved, electrons and quarks don't disappear just because they move around, even with something like controlled fusion. A Dyson sphere would be an extremely hot visible object, not some dark point.
Not my area, but I thought both were known to be incomplete? Q because it presumes a flat spacetime; R because it predicts the formation of singularities that the maths used to develop it assume don't exist?
It wasn't meant to be descriptive, just that there is a gap between theories.
And one theory is there is some 'unknown' type of matter, and they called it dark.
What I meant to write is—
> Dark matter isn't merely non-luminous matter.