You shouldn't consider the number of actual kill events. You should consider the number of events in general. Eventually a meteor will "get lucky" and take out a major city. Any such event, anywhere in the world, will have world-wide impact (no pun intended). It would be wise to prepare a solution, if we have the technology (which we do).
What your suggesting by not preparing is like the insurance question: "I've never been sick, been in an accident or had my home damaged; so, why should I carry insurance?" You pay for the insurance policies because it's wise to do so. The economic fallout of not having insurance will literally ruin you, it could take decades to recover financially, should you actually have an incident.
But they do.
Medical fields, electronics and semiconductor industry, chemical industry all spend billions each year. Automotive industry income sheets read in the billions per quarter and they all have substantial R&D expenses. The cost to build and maintain power plants and infrastructure runs into billions per year, nationwide.
Research into lighting technology, not so much; but, without the supporting infrastructure there would be no electric lights to begin with. I referred to lighting and refrigeration specifically because those were two of the earliest technologies that drove buildout of electrical infrastructure and had the largest impact. (power plants were initially being built to support lighting -- all the other stuff followed).
"It's the best thing since professional golfers on 'ludes." -- Rick Obidiah