multiprocess getting states from stub-states

[tbostic32]

@tbostic32 had a thought that I am slightly paraphrasing:

Instead of having our crawler go to each stub state and load the page to record the dom, kick that off to another process and the main crawler process continues to iterate through its other tasks. Loading the page tends to take a while, and we don't crawl any part of the stub-states, so minimal information needs to be passed back and forth, making it seem like low-hanging fruit for a multi-process application.