Re: "Now figure out how you're going to get all of that into space."
Nobody is seriously proposing that earthbound industries be physically moved into orbit. That's as silly as expecting westbound American pioneers to have brought in their covered wagons all the wood and bricks required to build their houses when they reached their new home.
What is possible, however, is putting just enough tools in space to begin processing raw materials already there. Land a little automated smelter on a nickel/iron asteroid, and let it start drilling and processing, using solar power. Takes a long time, but you start getting metal out of it to make more things. You want organics? Do roughly the same thing with a carbonaceous chondrite. Eventually, you get to the point where you can build larger industries, based on local materials.
Re: "Even when you've got it, now figure out how you're going to get it back down to the ground." That's easy in comparison. Solar sail for orbital changes, and a chunk of unprocessed nickel/iron as a heat shield, and a target (and payload) that doesn't mind a little thumping.
The critical thing here is time. It takes time to prospect extraterrestrial resources, time to process them, time to move them, time to create larger things with them (such as solar power satellites with a significant power output). Most of these activities would have to be automated, just because what humans in space don't have is time to hang about.
If we started taking XT resources seriously today, would we have enough time to forestall coming shortages? I doubt it: I think we've wasted the last fifty years arguing, and it's too late: when we suddenly need XT power and materials, they'll be fifty years away. But that doesn't mean that it's foolish to try.
