I've received your sequencer code on the dev list, and will take a look at it. I wasn't sure whether it'd be better to have multiple sequencers or a single sequencer for the different kinds of MS Office files.
Glad you're finding the project interesting. As for other activities, take a look at some of the smaller issues and feel free to submit patches (you could attach them to the JIRA issue for someone to review). I'll also contact you on email to talk about the specifics.
I'm going to tackle Jira issue DNA-76, which proposes providing a sequencing context to sequencers, and which also now incorporates your concern about MIME types. MIME type determination would be handled by the sequencer service rather than by a standalone sequencer. It does seem like we need to develop some sort of framework similar to the sequencer framework that allows for contributing processors that determine MIME type, names, etc., that would be populated within the proposed sequencing context.
That's definately a solution, however it must be clearly stated what should be handled by sequencers and what should be handled by the sequencing engine itself (or the processing engine).