There is a big big demand to get an exchange in which assets are not steal able. Neither by the people running the exchange neither by hackers. The single point of failure hurt a lot of people. There is no exchange I am currently trusting. Mt. Gox, Poloniex, Mintpal and other have proven again and again there will also be a point of failure. In the end it doesn't matter if a Government, an admin or an Hacker creates this situation where assets are gone.
You are absolutely right in emphasizing the importance of a blockchain 2 app, you are also absolutely right that the devs should forget all meaningless features and focus on the blockcain 2 function and it would be great if you could get this message through to the DEVs. Personally I think a decentralized market place, like a next decentralized Ebay that uses vericoin would be significantly bigger than a decentralized exchange, plus as users would trade with vericoin in the decentralized market place the demand for vericoin and consequently the price would sky rocket. Having said that, obviously the decentralized exchange would be great as well. Basically, any applications would be great that can be labelled as a blockhain 2 application.
Step:
1.) Create a base the exchange could operate on. Possibly a unique Blockchain 2.0 ? I am not sure if a decentralized exchange could even operate on the current blockchain.
2.) Create the exchange. Just a simple test version integrated into the Vericoin Wallet to find problems.
3.) Make the exchange accessible over the browser. Create an Veri-exchange only Wallet
4.) Add Altcoins
That's the main issue, how to achieve step 1.
For example Bitcoin use the OP_RETURN output in the transaction so you could say the bitcoin blockchain supports smart contracts in that sense. It seems to me the VeriCoin raw transaction could have OP_RETURN output as well so it is possible to build blockchain 2 apps on the top of that. At least my understanding was that the bitcoin devs suggest to use the OP_RETURN output and build blockchain 2 features top on that, but I could be wrong. Other projects like Viacoin take a different route and uses existing infrastructure like Counterparty which make sense too. I am sure that will work for them. If I would be the dev I would take the Ethereum route so it would be a contract entity and that could create messages instead of the transaction or collection of the transaction is the contract. In this case I guess you would need a ledger for the contracts, but wouldn't have to mess with the blockchain.
Either way lots of development, but I am sure the talented VeriCoin devs can manage that. If they need help the community could put together a 60-80K fund to hire peoples.