Professor Gregory Mandel has some excellent comments on KSR on Dennis Crouch's Patently-O site.

Though the Supreme Court’s unanimous reversal in [KSR v. Teleflex] contains some harsh words for the [Circuit’s] teaching, suggestion, or motivation (TSM) test, the decision itself appears to leave the TSM requirement roughly intact. Justice Kennedy’s opinion emphatically rejects an 'explicit' TSM test—one that would require explicit prior art teachings in order to combine given references in the obviousness analysis.

...the decision appears to essentially recreate the 'implicit' or 'flexible' TSM test—one that would allow implicit suggestions, such as the nature of the problem, to provide the requisite motivation to combine. It is this implicit TSM requirement that represented current [Circuit] doctrine, pursuant to several decisions published after certiorari was granted in KSR.

...the Supreme Court even indicates that the [Circuit] may have gotten it right in these post-cert cases. The KSR opinion is more a critique of the Circuit's application of the obviousness (and TSM) standard to the specific facts in KSR than a critique of the need to rigorously (and expansively) evaluate what would lead a PHOSITA to combine certain references in the obviousness analysis.

Yes, KSR was not quite the "smackdown" of the Federal Circuit that patent critics have mistakenly celebrated, more-or-less because they wanted the decision to discredit the Federal Circuit for resolving the issue of software patentability a decade ago.

Mandel also discusses "hindsight bias," which patent critics don't seem to grasp when they put on their tin-foil hats, come up with their own idiosyncratic definition of "non-obvious" and play amateur patent agent.

… a significant flaw in the KSR opinion is that it appears to misunderstand the hindsight bias and does not take it as seriously as is warranted. The hindsight concern... was the basis for the TSM requirement in the first instance. The Court states that a "factfinder should be aware, of course, of the distortion caused by the hindsight bias and must be cautious of arguments reliant upon ex post reasoning." The hindsight bias, however, does not refer to ex post reasoning, but to the effect that knowledge of the invention itself has on one’s ability to judge whether it was obvious when it was achieved.

Another issue discussed by Mandel is how the Patent Clause fits into the picture, a significant issue since patent critics have made various presumptions about "to promote the progress" and how the phrase relates to patent policy. One bizarre position against patents argues that each patent should be evaluated against the Patent Clause, rather than the patent system or aggregate effect of patents-.

The opinion in KSR explicitly notes the concern that granting patents on obvious advances might stifle progress in the useful arts, and cites the Constitution for this proposition. This statement could have significant repercussions in the context of patent thicket and anti-commons arguments, though I would hardly expect it to be actually applied to deny a patent on a constitutional basis as not promoting progress.

KSR may not be the radical, breakthrough that many hoped for, but very few things are that shape and influence the Internet economy.