The answer to your question about converting reflectance back to radiance can either be a simple one or a hard one, depending on what you're trying to achieve with your comparison. If you just want to do a very general comparison to radiance values of some other dataset or something, such that you just want to see if the numbers are reasonable, peak in the right place, etc., then you can probably get away with something pretty simple. Just take the equation for apparent reflectance and solve it for radiance. Your resulting radiances will not be technically correct of course, but might be close enough for your purposes. As luck would have it, another poster on another thread posted a link to a nice definition of apparent reflectance:
http://edcsns17.cr.usgs.gov/eo1/faq.php?id=21
Now, if you want a rigorous conversion that considers things like atmospheric gases, that will be considerably harder. You'd have to know all the steps that were done to create the reflectance in the first place, then reverse those - the modelling of atmospheric gases, etc. I don't remember what kind of outputs FLAASH gives you, but this seems likely to be very difficult to do and not likely to yield a lot of good results, but that depends on your goals.