Professor C.A.R. Hoare and other researchers have put together a list of challenges to computer science research for the years to come:
Computing needs a Grand Challenge
The Grand Challenges for computer science
Most likely, they are going to accomplish their goals.
In the meantime, what about putting Grand Challenges for software development, today?
Grand Challenge No. 1:
Turn mediocre software to a quality standard worthy of its price. Delight software consumers with software of exceptional quality, on time, within budget and complete.