The challenge of building large-scale complex systems often gets caught between purist visions that never get off the ground and seemingly pragmatic random-walk tinkering that slowly grinds to a halt via diminishing returns. If you're ambitious -- and I hope you are for the sake of my own future gig-flow -- sooner or later, whether you are an engineer, designer, project manager, CTO, CMO, VP of Sales, or CEO, you'll run into this challenge. How do you break out of this rock-and-a-hard place impasse?