The file extensions aren't always the best way of finding out things. File extensions are fluid and there's never been much in the way of standardisation, or at least not back in the DOS days. If you look at FilExt, for example, there's a fair bit of double up.
You'd be better off running the files through a tool like TrID/32 - File Identifier v2.10 - (C) 2003-11 By M.Pontello which does a good job of recognising files by their content rather than their file extension. It's not foolproof but can identify a few thousand different file types.
I used to do a lot of development on DOS back in the day. If you want to contact me off list, bruce dot axtens at gmail dot com, I can help identify the files and perhaps cook up a mechanism to extract the data.