He's right though, even people well versed in the technical aspects of Bitcoin have a hard time predicting all the various scenarios that can go wrong with proof of stake, let alone a full blown eMunie phonebook of variables.  You then have people like Kushti and Buterin trying to mathematically prove that proof of stake can rival or beat proof of work's security model in post's like this:

http://blog.ethereum.org/2014/11/25/proof-stake-learned-love-weak-subjectivity/

Since they have such a hard time quantifying the security model of simpler concepts like PoW vs PoS, how many decades will it take to quantify the security model of emunie?  Even if you release the perfect system, it might sit on the shelf for several years before anyone will touch it.