The truth is, it's not a paradox, because we don't really know quite what to expect at large distances. Very early in the life of the universe, there was probably an FTL inflation. The period when that occurred is believed to be from 10
-36 to 10
-32 seconds after creation. That is an unbelievably brief period, of unspeakably high energy, and how much clumpiness would propagate during that time depends very crucially on the specific masses of particles that have never been seen.
So, as I said in a post elsewhere, this is in an area where there are probably not so much "paradoxes" as simply "unknowns."
The Wikipedia article, and how Inflation accounts for the [mostly] large scale uniformity, some other things, and where it has problems, is decent: http://en.wikipedia.org/wiki/Inflation_(cosmology)