big cleanup

[nhoffman]

Attempt a significant simplification of the pipeline starting with the output of guppy classify

[jgolob]

I have a few ideas to simplify things (as in the branch pipeline). I've also confirmed the recompiled guppy CAN tolerate multithreading even when outputting to an SQLITE database.

[nhoffman]

@jgolob - where's the best place to look at your suggestions?

[jgolob]

I'll make a pull request.