Finding planets around distant stars is tough enough, but to find a virtual earth twin 1400 light years away is phenomenal! How much light does that bad boy reflect, anyhow? Not enough to invest in “extraterrestrial solar power”, I am sure.
So, how does NASA find these needles in haystacks? And, can’t this same technique be used as early detection to find killer asteroids? (I’m sure these asteroids give off just as much light.) 
They aren’t using light to find these, at least not light reflected off of the planet.
The technique is to measure perturbations in the star’s motion caused by the orbiting planet.
ETA: Here’s a recent thread.
I think they also detect the decrease in light from the star as the planet moves between the star and the Earth - a method that also won’t help with asteroids.
You could in principle find asteroids this way, but it’d be a huge gamble. Any given planet has about a 1 in 100 chance of being detectable via occultation, and we deal with that by looking at tens of thousands of candidate stars at once. That’s great if we’re just building up statistics on a whole bunch of other solar systems, but a 1 in 100 chance of detecting a killer asteroid isn’t too comforting.
Finding potentially hazardous objects (PHO) which cross Earth orbit and may eventually impact the Earth isn’t actually very difficult, at least in concept. A satellite observatory (or small constellation of them) orbiting somewhere between the orbit of Venus and Earth would be able to detect viritually all Near Earth Objects that are large enough to pose a hazard very quickly, because of the reflection of solar radiation off of their faces and movement against the stellar background. The PHOs we have difficulty finding are those that are generally inside of Earth orbit such that we rarely see the Sol-facing side directly.
A single satellite to perform this mission could be launched and deployed for somewhat less than US$2B (probably less depending on the use of hertiage systems and size of objects to be searched for) and multiple satellites could be roughly half that or a little more (assuming launches with the D-IV and Delta Cryogenic Second Stage, or Atlas V and Centaur). The total cost of, say, a five satellite system could be in the US$6B to US$8B range and detect virtually anything 50 m or larger in longest aspect, with sufficient fidelity in the orbital elements to predict the trajectory with high precision for years or even decades into the future.
Of course, then if we discovered a threatening PHO, we’d have to do something about it…
Stranger
That would work for the short-period objects, but wouldn’t be much improvement over what we’ve got now for long-period or open-orbit objects (think comets). I’m not sure what percentage of threats that would represent.
The most likely threats are those that are in orbits relatively near Earth, and especially crossing Earth orbit or readily perturbed by the gravity of the Earth-Moon system. Very long period objects from teh Oort Cloud or Kuiper Belt, while potentially posing a hazard of greater consequence due to their relative velocity and difficulty to observe or potentially intercept with a hypothetical deflection system, are so rare as to be almost dismissed, as can extrasolar objects which we would have essentially no hope of detecting, much less deflecting with any conventional technology. The biggest bang for the buck in terms of mitigating impact of threatening objects is to identify those which cross Earth orbit with the greatest frequency and then figuring out what is needed to deflect those which have a high potential for impact.
Stranger