There's a more meta-level discussion on meiosis, which is that there is an advantage for genes which can get themselves selected in meiosis more often would prosper. This is a prediction of the biologist Robert Trivers, working with Bill Hamilton in the late '60s, known also as the 'selfish gene'. In fact the reason the selfish gene theory was so successful was that selfish genes - individual genes which propagated to offspring more often than 50% of the time. The Dawkins book 'The Selfish Gene' discusses this whole idea and its ramifications in great detail.
So at different times in evolutionary history, its hard to imagine that meiosis was completely 'fair' (50/50% change of any segment of DNA being chosen). In fact its pretty clear from genetic analysis now that some longer segments of DNA tend to be transmitted in one piece, though overall its still felt that the propagation if 50% likely for any given segment.
Over the long run, its been shown, mostly through game theory, that honesty is the best and most profitable policy. The most famous of these experiments was the Prisoner's dilemma competition where all comers have been invited to submit computer programs which can either betray or cooperate. Overall cooperative programs with some safegaurds which punished cheaters win overall. Over time its been shown that systems tend to defend themselves from cheaters and that cooperation and 'fairness' do much better than mobs of cheaters running amok.
Cheaters and unfairness still continue to happen though - systems which are fair tend to invite in a small number of cheaters. Say what you will about the income tax as a system! there is variance over time as to how 'fair' a system can be.
Dawkins has even said he sometimes thinks he should have titled the 'selfish gene' something else instead: 'Nice Guys Finish First'. I doubt that the book would quite have done as well though... we all should be fair, but it doesn't seem to invite book sales.