Regarding the code itself, I don't see any challenges.
Most of the project is mostly taking the existing libzerocoin code that exists and converts it into an efficient solidity code, and pretty much serialize the outputs that come from mint transactions and the other transactions that occur in libzerocoin and making them work in the context of solidity.
The biggest challenge he sees is the inefficiency of it at the moment because of the amount of time it takes to perform heavy computational tasks.
In Etherum, there is a certain amount degree of gas that can be performed per block and verifying a ZCoin transaction will take up a considerable amount of a single block, and as a result that would be the biggest hurdle to actual implementation.
Once Etherum can scale to a certain degree he will definitely forsee this beinging a very advantageous to the Etherum platform.