ELO has been tried, tested and rings true for many years now, so it's always a viable solution, regardless of the format of a ranking system. The only problem that ELO has is that it needs many players involved in order to self-correct as designed. Making an ELO ranking system within a game with (apparently) a big number of bots involved is problematic, as it will highlight the real players' skills more than what they are in reality (assuming the bots are bots and play sub-optimal in comparison with real people in games that have a randomness element to them).
My 2 cents on the apparent discussion.