"A method of rendering a 3D animation in real time on a computer monitor is patentable subject matter." Just because you don't understand the subject matter does not make it patentable. Your example is a really bad example because it takes 2 coordinate transformations, one linear interpolation and 2 comparisons to go from geometry to pixels. From that point on it is just copying them from video memory to a screen. To anybody with a math background these are a really old hat and just because two specific matrices does not mean it is patatable. Computer graphics are a really bad filed for patetens because you are either in the field of basic math or trying to replicate physics.
Although I don't think that patents are inherently bad, the current burden of proof for software patents are inherently low. As being listed as inventor on two patents and having seen a few others, you can not claim that I totally do not have a clue how the system works. How I and few of my colleagues came to being "inventors" was, we just did out daily work of solving mediocre to boring problems and then someone comes along and is awed in his naivety and pushes us write a few pages about the "invention". This goes by lawyers and a few months later you are an inventor on the patent. To this day I have never invented anything, I just look around me and into other fields and am quite good at applying new ideas to old problems.
Maybe the LZW patents and similar groundbreaking algorithms may be worthy of a patent, but the current status quo you can throw out 95% of the patents, because any person with a modest intellect will come up with a similar solution to the given problem. The only difference is that patent holder had the problem first.