For chemical couplings, not all the nucleotides have same coupling efficiency so obviously you get a bias in oligonucleotides where you ask for 'N'.  Usually G and T have slightly higher rate of coupling. 

it is depended on the synthesis cycle being used, for high %G oligo, the manufacturer usually do capping-oxidation to cleave the N phosphorylation on G, the coupling time is also an important factor of this.

Thanks a lot for both of you - do you have any references to support this?