That the answer could be "not very well" (and we haven't heard otherwise) is what's most troubling to me.
It's say it's worse. I followed up on the general impression they were proposing to do this all with fingersticks, e.g. this citation from Wikipedia's introductory paragraph published in 2013: http://www.wsj.com/news/articles/SB1000142412788732412300457...
Based on my general knowledge of biology and medicine, I'd expect there would be a whole bunch of things you can't test that way due to general contamination from all the normal cells and their membranes that get punctured in a finger stick. Lipid panels, for example. In fact, the large venous draws normally taken, at least for the first tube, could be in part to swamp the contamination that comes from getting the needle into the vein in the first place.
Per other discussions, they're now using their own technology only for HSV-1 tests, which are PCR based and therefore require only a tiny sample, and the above sort of contamination would not be an issue.
Theranos uses non-venous blood for their proprietary tests (currently only 1 test) and rest 99.9% (figuratively speaking) tests uses venous blood in traditional diagnostic equipment.
Back to the same question - If almost all of their tests are done using traditional equipments, then why $9B valuation?