"Ever" is a pretty strong word, and I think he's going to look pretty silly in a century or so. For example, here's an estimate of travel costs with boron fusion rockets:
http://nextbigfuture.com/2007/11/fusion-propulsion-if-bussar...If either Bussard's polywell fusion or focus fusion turn out to work, that'll be achievable within a couple decades. As Moore's Law continues and we get better at simulating plasma, it's not that unlikely that some form of fusion will work out.
There are a lot of possibilities for non-rocket launch, including various space-elevator-like schemes, laser launch, and mass drivers. Even without fusion, thorium fission could provide plenty of power.
It'd be pretty expensive and slow to travel to another star with fusion...but eventually, with large solar panels in close orbit around the sun, we'll have an awful lot of energy to play with, and just maybe we'll figure out efficient laser or microwave power transmission sometime in the next thousand years.
On the other hand, maybe we'll just colonize the Oort Cloud and gradually migrate to other stars over the next million years or so without really trying.
(And, not that I'm holding my breath for this one, but if Woodward's right about the Mach effect we'll get to other stars pretty quickly.)
As for the reasons...the resources of the solar system are millions of times what's available on Earth. Once launch is cheap it'll be a no-brainer to start mining the asteroids.