Most clones so far have been of games which don't rely on a large amount of varying art assets, and where the levels can be generated programmatically with a simple algorithm.
There are a number of properties of a game that would make it considerably more difficult to clone, and whilst higher production costs are a factor, I don't think there's a direct relationship between the development cost of a game and its cloneability. Creative developers should consider what they can do - relatively inexpensively - to give themselves a competitive advantage over uncreative developers seeking to piggyback their work.
Here are some of those properties off the top of my head:
- reliance on hand-designed levels, so that the player feels like they are solving a problem that another person has designed for them.
- reliance on a large amount of high-quality artwork (like Huntsville).
- reliance on a 'black box' simulation where it isn't immediately obvious how to copy the functionality (e.g. a sports management game).
- inclusion of a well-written, well-presented story (like Monkey Island).
- inclusion of strong characters (and perhaps even their development).
- an online community providing social play (like Puzzle Pirates).