IMHO, although the court would likely have found a use of a knockdown assay to have been a sufficiently descriptive method of modulating the protein, that does not end the analysis. First, patent protection in such a situation would likely have been limited solely to the use of a knockdown assay to modulate the protein. Second -- no offense intended, of course -- but apparently the use of a knockdown assay is so obvious that it took some fellow on the Internet who read a blog post only a few minutes to think to do that. Such "obviousness" would likely cause even the limited patent to be invalidated.
Of course, there's still considerable room in the law, particularly patent law, to "game" things, but this ruling is at least a step in the right direction.
Your reply makes me hopeful that this decision will have a greater impact than I had hoped.