I think it would be interesting to study the probabilities of a successful front run attack in the two-stage commit and reveal model. The attackers need to put in the commitments to their front run transactions that are built based on guesses of what it will take to front run a potential market/limit order. They put in these commitments (and pay a small transaction fee for each commitment to a transaction) in every block in which a market transaction commitment has been submitted (or a subset of these depending on the state of the order book in the markets they are targeting and how profitable they think their guess will be). Then in the next block when they can see the various reveals, they need to decide which of their transaction commitments (if any) to reveal in that block to successfully front run and make a profit (or at least cut their losses from the commitment transaction fees). The probability can be further reduced by half (or more if they are competing with other front running attackers) if the order in which the revealed market/limit orders are processed is further randomized (say using the random number generated thus far by the addition of the random number, which was committed to in the previous round, revealed by the current block-producing delegate).
Depending on the transaction fee imposed for the commitments to market/limit orders (obviously it cannot be too high or else users will find it unacceptable) and the state (the liquidity) of the order books in the relevant markets, the probability of an attacker having a successful front running transaction committed in the previous block could be so low that the entire attempt would be unprofitable on average. In which case, no rational actor would even attempt to front run.
My guess, however, is that given the typical state of the order books and the highest acceptable transaction fee for market/limit order commitments, it will be profitable enough for front runners to pull this off. In which case, we are worse off than just having the front running proceeds go to the shareholders through enforced YGWYAF orders.
Too bad.