> Wouldn't we be better off making sure we can reliably find them first?

People are already are working on that. For example, there's the NEOWISE project[0], which used the Wide-field Infrared Survey Explorer to look for potentially dangerous astroids. And there is a ongoing effort for a new space mission, NEOCAM[1], specifically designed to do this same thing and find more objects.

[0] http://neo.jpl.nasa.gov/programs/neowise.html

[1] https://en.wikipedia.org/wiki/Near-Earth_Object_Camera

They propose additions use cases for the array in their papers. These include a LIDAR based detection system. I'm on my phone right now so I can't link the appropriate one. Let me know if you are interested and I can find it later.