Greetings -
In a completely-futile-give-up-now attempt to keep up with Gary, hot on the heals of the latest addition to the GOF options in MARK, I've revised and just upload a major revision to the GOF chapter (chapter 5). It covers Gary's newest GOF approach (the median c-hat method), as well as several other things.
The good news about the median approach, besides Gary's determination that it 'seems to work pretty well' is that it works with most (if not all) models in MARK - exclusive of individual covariates.
The bad news is that if you run a reasonably large 'simulation space' (which will make sense to you once you've downloaded and read the chapter), your computer will weep unconsolably. While I don't have the hottest hardware on my desk, even the venerable (and completely overused) dipper data set took a fair bit of time (about an hour) to get the estimate of c-hat.
But, I'll trade that off against the frustrations of not being able to get an estimate of fit for the fancy models the 'smart folks' keep coming up with.
Relevant chapters (and the updated 'single file' version of the complete book) can be found at
http://www.phidot.org/software/mark/docs/book/
Enjoy.