But will it realy make a difference in the long term ? I mean, if we take a genome and look at it mutate for a long time. Will not every nucleotide position, after several mutations, have a 50% chance of being AT or GC, and so when we consider the total ratio of the genome, we should expect to find approximately a 50/50 combination of AT and GC in it ?
Well, that depends on exactly what's going on. If we have a scenario as in "case 2" in my previous post, then no. In that case the AT content will have an equal chance of taking on any value. Note that this
does agree with your intuition that "every nucleotide position, after several mutations, [will] have a 50% chance of being AT or GC" ---
but only because 50% is the average of the numbers between 0% and 100%. 50% would therefore be the "expected value" only in the statistical sense of being the mean average outcome, not in the sense that we should particularly expect that the AT content should have that value rather than any other.