Um--this is a pillar of the present argument of evolution by variation and natural selection. The point being that--contrary to creationist's largely unstated assumptions that selection occurs over a uniform continuous distribution of evolved entities--selection occurs over a distribution with a central tendency. It is this central tendency that is the sculptor's knife, taken unwarrantedly by creationists as evidence of divine intervention, design, or irreducible complexity.
I do not dispute your assertion of random selection from a population with a central tendency. My assertion really has little to do with selection at all. What I suspect will undo randomness is information theory – particularly the discovery of algorithms and symbols. This is described better in the links on my post 103.
IOW, I suspect these researchers will conclude that the mutations presumed to fuel natural selection actually were more likely to result from information content - algorithms, including the use of symbols (self/non self – friend/foe – etc.) The process (and the result) could not be called random.