hi AK. maybe consider blogging about this? this is a well written but superficial analysis. scott aaronson insists on his blog that a proof should explain why it succeeds against "known barriers" eg razborov/rudich Natural Proofs. but this is really an optional requirement of a proof. moreover the actual barrier to "natural proofs" is very subtle and basically insists that a proof, if it exists, should have a certain "intrinsic complexity" in its constructions, and that many such constructions in the literature for class separations do not have this "intrinsic complexity". but new researchers are just suggesting that this only requires some new "intrinsic complexity" function that hasnt been seen before; but that such functions do exist, they just dont seem to be used in proofs that we know of. eminent authorities in the field such as Lipton have argued that the near 20yr old Natural Proofs may be overdramatized as a real barrier.
here is liptons blog on the subj: http://rjlipton.wordpress.com/2009/03/25/whos-afraid-of-natu...
here is chow in AMS:
"Nevertheless, it is my personal opinion that the optimistic approach is the right one; that is, the Razborov–Rudich result should be regarded as a hint, and not a barrier, to separating complex- ity classes. The only real barrier is our lack of imagination."