The Supreme Court rejected any specific test as dispositive for patentability. The Court of Appeals for the Federal Circuit [1] had advanced a "machine or transformation" test as the standard for abstraction and unpatentability. That test would have required that a patent specify a specific machine or a transformation of a form or state of matter somehow in order to be sufficiently concrete. It was a vague test that would have allowed almost anything not as awful as Bilsky's application, which essentially claimed various century old hedging strategies whenever implemented on a computer.
The Supreme Court ruled 9-0 that the test was an indicator of abstraction but not definitive. It also ruled 9-0 that Bilsky's patent was garbage.
The patentability of business methods in general was discussed in the opinions. The great John Paul Stevens [2] wrote the dissent in Diamond v. Diehr [3] thirty years prior explaining why and how software patents were terrible and should never be allowed, but only gained four votes. He wrote an opinion explaining again why software patents should not be allowed and gained four votes again in Bilsky in a court where every member but him had been replaced in the interim.
Unfortunately Stevens is retired (it's well earned -- he's 93) and the leading intellectual property expert on the court is now Breyer. Breyer wrote the dissent against effectively permanent copyright terms in Eldred v. Ashcroft. Sotomayor and Ginsberg have also been in the Stevens camp on patentability. New Justice Elena Kagan usually votes with those three. So we can hope for four votes for software freedom once again.
There were also four votes in Bilsky for patenting software, though not with any visible enthusiasm. They were Roberts, Alito, Kennedy, and Thomas.
Finally Scalia refused to join just one opinion and split between the two without addressing the serious ongoing questions of patentable subject matter. So the final tally was 4-1-4.
The CAFC split 5-5 on the Alice [4] case the Supreme Court just decided to hear. The split was similar to the Supreme Court in Bilsky and did not resolve any simmering issues because of the even split.
Note that even the Google and EFF briefs on the Alice case didn't suggest outright abolition of software patents at the CAFC. Justice Stevens may have seemed like a lonely crusader by 2010. It may be that the enemies of our industry have gained so much power in Washington and the patent bar that the best we can hope for is a slow, slow rollback. Certainly there are a lot of people making billions without the trouble of actually building anything who can spend their efforts to keep the racket going. Without the need to make things, they can lobby full time to continue getting rich at the expense of innovators and startups and the public through software patents.
The most likely result here is a 4-1-4 split again, unless a new way to appeal to Scalia can be found. Real progress from the Supreme Court probably requires a new justice; the current pattern indicates that pragmatic Democratic appointee is much more likely to abolish software patents and a movement Republican appointee is most likely to definitively endorse them, but individual justices can always assert their own visions once appointed.
[0] http://en.wikipedia.org/wiki/Bilski_v._Kappos
[1] http://en.wikipedia.org/wiki/John_Paul_Stevens
[2] http://en.wikipedia.org/wiki/United_States_Court_of_Appeals_...