As some of you may know, I have recently made the move from optics, lasers, and fun, to... um... surface science, chemistry and, well other kinds of fun. As far as I am concerned, the difficult thing about surfaces is figuring out what is going on. Everything that you are interested in is happening within one layer of atoms, and that presents some challenges. One technique that we worked with very early is called ellipsometry. Ellipsometry has a very simple recipe: take light with a very well defined polarization, reflect that light off the surface, and measure the polarization of the reflected light. You can use the change in polarization to determine what is on the surface.
In practice, however, ellipsometry suffers from a significant challenge: getting polarized light anywhere near the surfaces we want to understand. In complete ignorance and with the confidence that entails, my response was "bugger this, just run the light through an optical fiber." The ellipsometry people I suggested this to stared at me as if I had grown a nipple on my forehead.