Lang has several different models of the same kinds of subjects, which Satoshi doesn't.
Plus, most of Lang's newer stuff is either abstract (Squaring the Circle, 8.8.4 Resch, etc.) or box-pleated, like most of his insects. The only box-pleats Satoshi has are the Ryujins, Mammoth, Shishigami and Caribou 2.2--if only partially--and a few others. There's only so much that can be done in the way of purely representative designs that can be dome without exhausting techniques. I also noticed that many of Satoshi's designs are based on video-game characters. While there are many of these (and his renditions are amazing!), they're obviously very difficult to fold, and I wouldn't be surprised if Satoshi's "worn himself out" as far as design techniques go.
So really, it all boils down to the choice of subject material. Lang has chosen to pick from a much wider variety, even going so far as to do a few models of one subject. Satoshi, on the other hand, seems to pick and choose his subject from almost everywhere, but not doing much in any one area.
HankSimon wrote: I guess the next breakthrough we need is one that encourages diagramming...
I actually don't agree with this. While we all would like to see more diagrams, origami is an advancing art. Models these days are getting so complex as to be un-diagrammable. Look at Lang's Roosevelt Elk from ODS2 for instance. More of today's models being diagrammed would mean more of "measure-and-mark-these-20-points," which would take away from the purity of the art. And to me, nothing is more important than preserving the integrity of origami. Would Yoshizawa have been content to sit for hours with a ruler, carefully finding to-the-10,000th reference points?