I respectfully disagree :-)
I normally agree with you, Andrew, but I think you have got this one wrong. I am qualified to comment as a UK and European patent attorney, with 18 years experience of procurement with a little contentious work.
Before I start, I should say that you probably meant to qualify your statement “it's the norm for juries to decide patent disputes” with “in the US”. The statement is not true for most other jurisdictions.
I would argue that the fact that patent cases are tried by juries is one of the factors by which patent litigation in the US is as awful as it is (others include triple damages for wilful infringement and the fact that the winning party cannot claim their costs from the loser).
Trial by jury introduces a randomiser and makes it less likely that the ‘right’ party will win. Juries are a randomiser because they have to decide on issues that they don’t fully understand, or in some cases don’t understand at all. These issues include technical issues and legal ones. Some of this stuff is really hard to get to grips with, and the less you understand it the more likely you are to make the wrong decision. The process is randomised further because the parties can deselect potential jurors. If you have the weakest case, you deselect the smartest jurors (a US attorney told me that this a commonly used strategy).
Trial by jury increases costs. If you have to spell everything out in court in plain English to lay people (some of whom will not be the sharpest tool in the box), attorney time and court time goes up. This increases the costs for everyone. And you don’t get your costs paid by the loser when you win.
In the US, judges overseeing patent cases may not have any technical expertise. In other countries, they often do. In other countries, judges can appoint technical advisors to help them with the technical issues. This happens quite a lot in the UK.
In the US, judges overseeing patent cases may not have any experience of patent law. Experience counts for a lot with patents as with most complicated things, and it is only with experience or a very very sharp mind that you will be able to see the interplay between various (often individually tricky) issues. In other countries, patent cases are decided by specialist patent (or general IP) judges.
If your judge is experienced in patent cases and has technical expertise and/or an advisor that does, they are going to come to the ‘right’ decision more often than a jury. If the court is going to come to the right decision, the wrong party will settle. If the wrong party might win, the case will proceed further and is more likely to go all the way to trial. If the wrong party is bigger than the right party, bullying through (ab)use of procedure is almost inevitable.
The thing that bugs me the most is the way the various ‘features’ of US patent litigation work together such it is cheaper to settle than it is to win at trial. Non-infringers and infringers of invalid patents thus pay patentees when they shouldn’t have to. The balance really does swing in the favour of patentees, especially against SMEs. Thus, patent trolls are very successful. They would be less successful if patent trials were not jury trials. They would hardly succeed at all if there were no triple damages and the loser paid the winner’s costs.
And this is not even going into the issue of calculating damages …