It was never as good with the patent issues especially when Google were involved it never felt balanced (with SCO there really wasn't any need for balance or bias as SCO were so far out there you could report it straight and still have clear bad guys).
It was fine for Oracle v. Google as reporting that one straight provided an obvious bad guy, too.
And Samsung v. Apple I also provides many obvious bad guys if you follow the case and understand the industry and read the patents. Apple's claims are garbage but Samsung had so much contempt for the process that they let their lawyers utterly fail to make the case. The judge lost control of the case and ended up with vague and useless jury instructions and a self-contradictory verdict (parts had to be retried recently). The jury foreman admitted to bias and misconduct in public interviews. The PTO keeps gyrating back and forth in 'final' decisions about the validity and meaning of claims to patents at issue. I never saw coverage anywhere else but Groklaw about most of the pathetic failures in the case.
Let's hope the Apple ][ and Apple /// cases are handled better.
The Motorola v. Microsoft case, on the other hand, revealed bizarre bias on the part of Groklaw. Even when Microsoft was obviously right, Groklaw manufactured sympathy for Googlerola on every issue.
But that's not really the question. The essential thing about Groklaw was that they would recruit volunteer reporters to go to court and make extensive and informative reports about what really happened they very same day. Nobody else's reporting ever came close.
I'm not even sure about that one. Oracle are a bad guy (always) but I'm not sure Google weren't one too in that particular case. If you replay the situation in your head imagining MS in the Google role and Sun in the Oracle role would you have felt the same way?
I think the conclusion that API's are not copyrightable is pretty much preposterous from my understanding of copyright law. A decision that there was a 'fair use' right to copy API's for compatibility purposes would have been an interesting and probably valuable development although I'm not sure Dalvik would have applied here.
Groklaw's example API to explain what API's (adding numbers I think) was also applied when considering whether there was creativity in expression and correctly the conclusion was that there isn't in the case of adding but then Groklaw inappropriately extrapolated this view to all API's.
The court reporting could be useful though you are right. Florian Mueller (http://www.fosspatents.com/) attends some of the German ones and I find his reporting on the patent cases very good. Groklaw seemed to have taken the view that you couldn't believe a word he said but I never quite understood why (although it might have been the Oracle v. Google case as he didn't see it the Groklaw way and had at some point consulted for Oracle).