When people start talking about doing rewrites and refactoring I know i'll never see their game.
Not really. I talk about refactoring all the time and I still release every month (or two). If my code is still maintainable after 13 years of development, this is only due to refactoring. Refactoring is a great thing when it's done right... and when it's done *instead* of rewriting, which is always a disaster to any kind of software.
By the way, Fame has survived inventing of a new language AND a rewrite of that language. What is more, it has undergone a complete rewrite of the whole game in a new language. So rewriting is a bad thing, sure, but there are worse. I don't want to know what was that in this case.
//EDIT:
OK, that was a lie. I actually wanted to know. This little quote says it all:
It's going to be a while before the gameplay is anything resembling the complexity of old-Incursion -- and once it is, it will quickly skyrocket past that, since the whole point of the script language and next-gen engine is to allow a lot of d20 content and gameplay detail to be added rapidly in an easily maintainable manner. Apparently implementing and troubleshooting an entire Java-clone programming language complete with bytecode VM, IDE support, symbolic debugger, garbage collection and various game-specific language extensions takes a substantial amount of time to get working and stable.