Meteors large enough to detect are both incredibly rare, and are very difficult to deal with, and meteors too small to detect will hit you before you see them.
Not to mention that any detection will come with very large uncertainty bars. Suppose that you somehow discover that a ~500 KT meteor (Like the Chelyabinsk bolide - which was only ~20 meters across - good luck figuring out an accurate orbit for it) is an impact risk. The overwhelming majority of the Earth is covered with water. The overwhelming majority of the rest is... Not densely populated. And observations put it at a 0.5% chance of striking the Earth next year, with zero ability to predict which part of the Earth will be struck.
That's the most likely outcome of a detection program. Not '100% chance', not '90% chance', but a '0.5% chance of a rock that's scary only if you're in its immediate path'. Is that worth doing anything about?