We can surely expect several months of creaking infrastructure. The type of project that can provide a substantially better IS would typically be a multi-year development project. Given the hotness of bitcoin I would guess there is going to be fierce competition for being the first noticeably better exchange and with competition comes efficiency enough to possibly cut out the common delays.

The greatest barrier for success here is credibility and trust. Whoever aims to beat mtgox at their own game will need lots of transparency. Who is the owner? Who are on the board? How much funding do they have? How do you use their API's? How is your money insured once at their exchange? etc...

All these requirements pile up, you'll need a seriously big budget. Hundreds of millions of usd. If you try it with less you can at best take a small market share and then you will need to spend several years with small scale operations to slowly climb up towards meeting your goals of actually improving the bitcoin ecosystem.