As a part time professor I've seen a number of programs that had different formatting, comments, variable names, but could be mechanically shown to be equivalent (via a fairly trivial set of search and replace operations). There are [link|http://www.berkeley.edu/news/media/releases/97legacy/11_19_97b.html|programs] that have been developed that check for this.

So lets see how the MOSS metric works on the various "obfuscated regions" and then we can make an informed judgement.