The idea that the USofA could not launch multiple nukes on deep space trajectories on three months notice is ludicrous. So could Russia and I suspect that the PRC would toss a couple of their own as well.

This might just work, if the blasts are at the correct angle simultaneously. Glad nobody came up with the idea to move earth.

Could also be used to deliberately target earth with a comet.

The laser-based system could work on asteroids, but if time was off the essence explosives, preferably nuclear, could be more effective at a shorter time period.
A fleet of small space probes that detonate at close proximity to a asteroid, at the same side of the rock over and over again, could gradely push it on a different course.
Controlled blast at a distance would make better use of its energy than what a surface explosion would do. The atomic bomb that destroyed Hiroshima detonated 1,968 feet (600m) above the city. It was more efficient because of the distance that existed between center of explosion and ground.
The approach described in this article could be effective against comets. A comet is a so-called dirty snowball, many small and big rocks held together by frozen water. Is more fragile and less dense than what a rocky asteroid is. A explosion could break the comet apart, but it would most likely continued on the same course. Explosives against comets is probably a bad approach.

Might as well use this same tech to shift useful material to L1 or lunar orbit.

do the math,--moving a solid mass sphere of radius of 30km is ludicrously impossible to do for large distances, especially over short periods of time.
i did the math of trying to crash phobos or deimos into mars and the energy required---assuming 100% of all energy necessary to move the mass is efficiently produced by a 100% energy efficient process-----is ludicrous. even deimos, with a radius of 6+ kilometers and moving at 1.35 kilometers a second relative to mars is nearly impossible to move and significant distance closer to the martian surface, even over a 100 year period.
to theorize about asteroid deflection is one thing. however, if you take for granted that we have small tactical nukes that can be hoisted onto existing space rockets for TESTING on real asteroids, using 1 kiloton explosions for TESTS on real asteroids [ which are far more likely to hit earth than comets ]------------then you are actually looking into empirical testing. not theorizing.
the best candidate for 'experimenting' with deflecting asteroids (for a madman perhaps?)-----is 1999 an10 http://www.halfbakery.com/idea/killasteroid#1380738091

Cost-effective to save the planet really.

And just what keeps those reflectors facing the "right" way? Could this not also be fired at dirtside targets? How much of that 10 kilowatts would make it through our atmosphere? With the precision required, perhaps even human targets?

& how about a Lunar based remote Laser as well aside those Lasersats. Other problems with Lasersats:
1. Hacking 2. Use for war over Earth.
Otherwise Yes, mass produce & launch in HEO or cisliunar orbit mode.

If there's something on a collision course with Earth, cost effective = whatever it costs to stop it from hitting Earth.
Then just like Y2K there will be all the ninnies going on about all the hype over something that wasn't really a problem. Duh! It wasn't a problem because of all the effort and expense people put forth to ensure it wouldn't be one.
If nothing had been done, they would've sure thought it was a problem to wake up on January 1st to discover all their power and gas shut off because the computers thought it was the year 19,100.