Furthermore, if she's got patents on these processes, can she wield them against anyone who tries to implement them successfully?
In particular, it seems that the microfluidics stuff needs new inventions to make it work in practice. So presumably those aren't described in the patents. Although it wouldn't be the first time that management overlooked solutions provided by their own researchers.
> The specification [of the patent] shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same
It's the raison d'etre of patents. A time-limited monopoly in exchange for full documentation of the invention for the long term public good.
If the method would work in general, then yes, the inventor with the patent can wield them against anyone who implements that method - even if the inventor wasn't able to build the apparatus due to expense or difficulty; that's part of the original goal of patents, so that the inventor can fundraise for building the device without the funders just building it themselves after being convinced that it will work.
If the method is bullshit and can't be used to achieve the intended goal, then the patent can't be used against someone who succeeds, since they would succeed with a substantially different method that's not covered by the patent.
Whoever ends up liquidating Theranos will be in a position to make a call with licensing and enforcement, assuming, of course, they don't let the patents lapse (which is probably a shorter timeframe than a complicated bankruptcy will take).
Letting the patents lapse is a deliberate strategy sometimes (it's more a defensive rather than offensive approach, to block other actors from blocking you from your own technology).
It's also an alternative to peer-reviewed journals for "publishing" your ideas.