Originally Posted by ATDrake
NCX files are XML, so it should be pretty trivial to use XSLT to yield an XHTML version. The trickier part is auto-inserting it into the .opf so that it shows up correctly in the resulting .mobi file, but I'm sure someone can come up with a simple scripting solution.
Edit: Huh. Officially a Fanatic now.
One additional complication is that the epubs are usually divided internally into several files, usually based on chapters. On conversion to mobi, apparently, all these files are combined back into one. So the targets of the TOC would have to be adjusted both in the TOC itself and in the internals of the html file.