For our final criterion, I'm going to mention something that may seem silly, but it's still a necessity. The solution needs to be useful; it can't be so weak that it has no real applications. This is perhaps the biggest problem with the current guiding principles of the legal system; they are so weak now as to be nearly useless.
Intellectual Property law's basic guiding principle is to promote innovation and thereby good to society, at least in theory. It's a good principle as far as it goes, but it has the unfortunate problem that there are multiple good ways to do that. Both sides of a debate claim they are supported by this principle, and generally both sides are right in their own way. The principle doesn't have any usefulness. This means that the court cases are being decided less on the merits of the case and more on the particular biases of who the judge thinks is innovating. For instance, in the first round of the Napster vs. RIAA case, in my opinion Judge Patel essentially ruled for Big Money and against the Internet Punk Thieves... which may be a valid ruling (assuming that Napster also represents Internet Punk Thieves, which is also another discussion) but was not truly informed on the basis of promoting innovation, but rather on the basis of preserving the interests of the status quo. (Note I also would have ruled against Napster, just for entirely different reasons, so this is not sour grapes about a ruling by one of said Punk Thieves.)
So if I get done spinning a theory but that theory never says anything is right, wrong, or up to the society to just decide, it's still completely useless.