- It takes a very short time in comparison to the age of the galaxy for a single self-replicating probe to result in a descendant probe visit to every star. A few million years. This means that you cannot really argue plausibly that all civilizations fail to do this - it would take trivial effort for just one faction within just one advanced civilization to set this underway. It's quite clearly the case that we will do it when we can.
- There is a proposal I recall reading that suggests all colonization/self-replicated expansion is patchy in a sort of fractal way. You're always going to get sparse frontiers and empty real estate.
- The Great Filter. http://en.wikipedia.org/wiki/Great_Filter
- The Simulation Argument (Bostrom variant). http://en.wikipedia.org/wiki/Simulation_hypothesis
- The Drake Equation http://en.wikipedia.org/wiki/Drake_equation
- Where are the macrostructures? We don't see Dyson spheres or other evident macroscale engineering. Everything we can see so far looks natural. This is more troubling than an absence of detectable communications. There are any number of ways we can think of for interstellar communication networks to exist and yet be undetectable by us at the moment, and then there are no doubt all the ways that we can't think of.
- Agents responsible for the Fermi Paradox through suppression of nascent intelligence are a popular science fiction trope. First one to the top kills off the rest as they turn up. (Berzerkers, Wolves, etc). It makes for good plotting, but in reality this suffers from all the same issues as economic monopolies and monocultures - i.e. without some sustaining outside influence they are fragile and short-lived, prone to defection, being undercut, and out-innovated from a blind spot. So it's not plausible to suppose that we are randomly in a period of suppressive monopoly.