I tend to agree with you about jury trials, but having the “real” decisions made by specialist judges in the Federal Circuit has dangers too. Many legal scholars believe that the trend toward allowing more software and business method patents in the 1980s and 1990s occurred because of Federal Circuit judges who were too embedded in the patent bar, and that it was allowed to continue without enough review because with all patent appeals coming out of one circuit, there were no circuit splits (disagreements between circuit courts of appeal), which is the usual way cases reach the US Supreme Court.
In any case it would be difficult for the trial judge or the Federal Circuit to ignore jury decisions on factual questions like the ones that were decided here, unless the result was egregiously wrong. Legal questions like whether APIs are copyrightable will almost certainly be reviewed on appeal.
[added] The dangers of judges who are too specialized was actually recognized when the Federal Circuit was set up, so the court was also given jurisdiction over a few other areas, such as appeals from the Court of Federal Claims, to prevent it from being overly specialized. I haven’t seen recent figures, but I do believe that the docket is dominated by patent cases.
It's not just that Federal Circuit judges are embedded in the patent bar, but also that judges are quite intellectual and academic and it is in vogue among academic circles to treat intellectual property like real property. Real property has a very strong right of exclusion (you can't be on my land no matter what you're doing on it, whether it hurts me or not or benefits society or not).
Economists classify goods as being either rivalrous or non-rivalrous, and either excludable or non-excludable.
The ownership of IP is, in a sense, a private good (excludable and rivalrous), just like a car. Take patents - can buy and sell them, they're hard to steal, and only one person can have it.
However, the right to use (not own) IP is not a private good. While the original copyright or patent may be unique (therefore rivalrous) and easy to defend (thus excludable), expressions of it are not.
Non-excludable, non-rivalrous goods (the right to use IP) are a little harder to reason about.
Cars are excludable and rivalrous property based on the fact they are physical objects that are of a size easy to possess and it is very hard in physical terms for them to not be so.
Patents, on the other hand, are excludable and rivalrous only in the tautological sense. They are defined within law as excludable and rivalrous, so therefore they are excludable and rivalrous.
Patents could be redefined in numerous ways that are not excludable and rivalrous, such as including anyone else who can give reasonable proof that they came up with the idea at the same time or independently. Also, it would not be that great a stretch to limit their transferability, so that they cannot be treated as a normal asset.
In any case it would be difficult for the trial judge or the Federal Circuit to ignore jury decisions on factual questions like the ones that were decided here, unless the result was egregiously wrong. Legal questions like whether APIs are copyrightable will almost certainly be reviewed on appeal.
[added] The dangers of judges who are too specialized was actually recognized when the Federal Circuit was set up, so the court was also given jurisdiction over a few other areas, such as appeals from the Court of Federal Claims, to prevent it from being overly specialized. I haven’t seen recent figures, but I do believe that the docket is dominated by patent cases.