Well, to quibble if you are data mining a collection of HTML files you probably need to do some reverse engineering of the format before you actually dig for data.
No quibble at all - you may need to do some reverse engineering before you do the data mining, but they are two distinct operations, The first is directed to undestanding the file format, the second is directed to extracting data from the collection of files (e.g. by looking for clusters of data or correllations between features) rather than just interpreting a single file.
In the case of iStripper reverse enginerring the animation files by using a hex editor and looking for patterns in the data would not be acceptable (though it would be very hard to do). This is
in principle no different to using a hex editor to look at a an understand a simpler file. There is of course a big difference in practice but the rules are written to cover the general case and in practice that reverse engineering simple file is usually tolerated even though it may ***** a strict interpretation of the licence (partly because it is simply not worth persuing any such *****)