A DNA locus may have two (or more) variants (alleles), but there isn't one termed a main or default variant. In the example you cite, Myśliwska 2009, the only asymmetric distinction between the G and C alleles that I could see was in this passage:
The polymorphic region −174G>C of IL-6 encoding gene is implicated in
transcription of this cytokine. The G>C nucleotide substitution
This passage uses "substitution" (and "G>C" in −174G>C), which suggests that "G" is what is termed the ancestral variant, and "C" a later variant. The ancestral variant was the form that is believed to have been essentially the only form in some ancestral population; the later variant is a mutation (substitution in this example) which became established in part of a later occurring population.
Not all polymorphisms have their variants characterized in this way. Those that do usually have DNA sequence evidence showing that species other than humans have a homologous gene, and those genes appear to have only the "ancestral" variant. For the locus in the example you cite, the human IL6 -174 locus, a early paper by Fishman et al. (1998) says this:
Considerable interethnic variation in the frequencies of these
polymorphisms has been demonstrated (39), which is consistent with our
data for the -174C allele, which is considerably rarer in Gujarati
Indians and Afro-Caribbeans, compared with UK Caucasians. As all of
the primates examined were GG homozygotes, it is likely that this
allele is ancestral and that the C allele represents a relatively
recent change in the IL-6 5' flanking sequence.
This assessment seems to have become generally accepted. The rather even distribution of G and C variants in the table you show is likely due to the subjects being of European ancestry, as suggested by this from SNPedia:
It [IL6 -174] tends to be quite polymorphic in Caucasians, but Asian
and African populations are almost monomorphic (for the (G) allele).