[pothole]

I believe the consensus is that actual engine failures are in the low single digit percentage region. On the one hand, that's pretty bad if true compared to what you might expect from a properly engineered engine. On the other, think of it this way - if you ran 50 Boxsters in parallel, you'd probably only expect one of the engines to fail. In that context, if you only buy a single Boxster you'd have to be fairly lucky for it to be the one that dies.

The way people talk about it, it seems like you're really rolling the dice on a 50/50 chance of the engine going pop. But I don't think it's anything like that bad a risk.