The idea is no one really knows what the difficulty will be in December.
So you have various levels of X, and they require different contribution shares from each party.
To define those levels you gave in the example you would need to know the probabilities to start out with. But as you say no one knows those probabilities. And the different contracts would become (un)interesting as new information becomes available and so the community estimate goes up or down. The icbit Dec contract started out at 300M and it last trade was at 484M. This free trading of these future contracts allows the community to bet on the future difficulty without anyone having to guess the probability upfront. It allows the price to reflect news as they become available from the ASIC vendors. And if enough people participate, the efficient market hypothesis predicts that the market value should be a good estimate for the December price. In that case it would not only help miners to hedge but also give them this good estimate as free information. Right now the estimate is probably not very accurate because the volume is low. On the upside, this gives a good opportunity to anyone interested to bet on the difficulty and who think the difficulty will be lower or higher than currently traded.
The market can define those probabilities just like the icbit contract. It's just a different payout structure with the same concept. To figure out the true value of the icbit contract, you need to factor in those probabilities just the same. A single binary contract actually requires less calculations than a variable contract and is something that can more easily be guessed.
Since the payout is variable based on how far it is off, you could think it is more likely to be below the trading value, but still have a +EV trade by buying the contract rather than selling, if the payout is high enough with the upside, and the payout on the more likely end is lower.