I believe this viewpoint is shared by a number of people. Specifically, that a litigious NPE is, by colloquial definition, a "patent troll." There is a conversation to be had as to whether the practice is good or bad, or if they are acting more appropriately than their peer group, but I think the term fits in this case.
In general, the law allows you to sell your cause of action to someone else. For example, many investors who had suits against big banks over mortgage-backed securities sold their causes of action to funds that could bring them efficiently in one go. That's just property rights in action. At least in theory, it's more efficient to allow rights of action to be sold like property.
Now, maybe you think that these sorts of suits shouldn't be brought at all. There's arguments to be made for that. But in that case, I think you need to lump companies like Apple in with companies like these.
Simple explanation: if they make their money from suing people, it's a troll.
proceeds to take own advice
[1]: such as this one, but this wouldn't be necessary if folks didn't find it necessary to waste space dictating what's worth talking about.
So, I think there is a reasonable argument that if your actively doing research your not really a NPE. AMD and University's both make money from both obvious and non obvious patents but they have positive contributions where companies that only buy existing patents and sue don't contribute anything as the patents already exist.
We need to understand what we're fighting. Defining a patent troll as any NPE who files an infringement suit is clearly too broad, because the only way to prevent trolling under that definition is to shut down the patent system. As long as patents are issued, and can be bought and sold, it will be possible for NPEs to own them. Furthermore, independent inventors, of whom I believe there are still a few left, and whom I think we all believe the patent system should protect if it protects anyone, are generally NPEs.
This is actually _exactly_ how language works, and that's precisely what's under discussion here: whether the term patent troll is accurate.
There's a different discussion to be had (which rayiner is also alluding to in his comment) about whether this definition has negative side-effects (like being too broad), but being clear on the terms being used in a discussion is important.