The overall challenge to me, I think, will be graphics (models and world designs could take a while themselves) and AI. Right now I am working on my own program and I see most of the core programming being finished fairly quickly.
When I programmed in C and C++ I found attention to detail took up my time and instead of spending time improving effects etc I was debugging. I always felt that DB1 had potential to be quite proffesiional depending on the game but had a lot of limitations. I didn't see any with DBPro except perhaps the speed which we are assured will be improved.
As I said I was only worried by the fact nobody appeared to be considering a large game.