This question was the subject of a noisy debate over the years leading to NEAR, with strong proponents on both sides. With the NEAR result, the question is settled.
If we decide to deflect an asteroid we will need to know just where the center is, and we need to know something about its mechanical (compositional) strength.
The surface is strewn with very angular, large rocks - much more so than was expected. They've picked the most benign looking spot, but it has not turned out to be very benign.
Rubble pile asteroids are something we always figured would be out there, and suddenly, 'voila,' there was one in front of us.
The object has not been co-operative at all. It is an incredibly nasty place to land,