Even more important, planetary astronomers have never observed a large impact of the sort this thing would make. We all know how valuable watching the Shoemaker-Levy collision was; imagine what we could gain by watching a 1-km asteroid hit a terrestrial surface right near home!
No, not Earth, I'm not that crazy. Let's try to make this thing hit the moon! What astronomer wouldn't give his eyeteeth to watch that happen? What geologist wouldn't do the same to be standing on the moon (at a safe distance) when it happened?
My rough calculations suggest that to move this thing around within its current approach-distance error (100,000 km; the moon's orbit is well within this error patch), you'd have to change its velocity by about .1 m/s 30 years ahead of time (i.e., now). That's equivalent to performing 5*10^12 Joules of work on the 'roid. Delta-v (and thus work) required is inversely proportional to lead time (twice as much energy at 15-year lead, etc.)
Humans produce this much energy in electricity every 2 seconds. It's equal to 1.2 kilotons of TNT (but you'd need lots more nukes than that to move it, 'cause the conversion efficiency is low). Since it weighs about 1.5 million tonnes, a single 7.5-tonne mass moving at 20 km/s would be enough to change its velocity by the required .1 m/s.
Wow. Is it really that easy? All we have to do is launch a Saturn V at it? Somebody check my math.