I didn't see Meteor but in the film, Armeggedon, I think they had a better plan in regards to the effectiveness of such a weapon: to drill into the body and plant the bomb inside. In the film, the scientific idea was to use the nuke to split the asteroid in two. While it appears this may not be possible on an object the size of the one in the film, I think the concept would yield a better result than exploding a nuke either nearby or even just at impact on the surface.
That said, I'm not sure how practical it would be to try to put a bomb plus drilling equipment up into space and then still have to have it bore into the object. I think the idea of sending astronauts up to do the drilling is even more ludicrous.
Splitting it is absolutely the wrong thing to do. The thing is a falling rock does an awful lot of overkill. Energy spent digging a crater doesn't kill anyone--they were already dead anyway. The damage radius goes up at the third root of the energy, since the area goes at the square of the radius the result is the damage area goes at the 2/3 root of the energy.
Lets take the simple case, you split the threatening object into 8 equal parts. Each part has half the damage radius (third root of 8) and one quarter the damage area. There are 8 parts, though--8 x 1/4 = twice the total damage area. (The math works regardless of the number of pieces, I just chose a number that wouldn't cause any irrational numbers.)
The only reason I can see to go for a split is if you are limited in blast energy rather than total mass. A carefully placed set of charges could perhaps blast it into two halves that are receding fast enough to generate a miss--this gives about 10x the efficiency of your bombs that nudging it aside would do. However, it requires landing and drilling to place the bombs while nudging can be done by a bomb that's coming in like a bat out of hell. (And note the geometry of an intercept against a rock that's going to hit soon--it's coming down our throat. No way to put enough delta-v to land on anything less than an Orion.) You had also be pretty darn sure of yourself if you're going for a split--if it doesn't split cleanly you'll have a bunch of rubble with no chance of stopping it.
I have a hard time picturing a situation where you are blast limited rather than mass limited. (NASA says a dinosaur killer is 3 months out, they'll have their pick of the world's nuclear arsenals for the asking. The limiting factor will be boosters.) If the rock is so big that we don't have enough megatonnage it's going to be so big I doubt we can drill it for splitting.