Seth Levine makes some good points about the real lack of transferability of NOL's in an acquisition. I think it's a little bold, especially given the current deficit spending climate.
I think a more realistic step is to make the U.S. R&D Tax Credit permanent. In addition, it must be made more competitive with the rates and structures in UK, Canada, China and India. The U.S. IRC is the most complex in the world, but the R&D Tax Credit is a joke - how can a concept so easy be made so hard. Finally, unused R&D tax credits should accumulate and be fully transferable in an acquisition.