Being peer-reviewed does not mean that a claim is valid; it was never intended to mean that, or to be a stamp of approval on the merits or results of research.

Peer-review is just intended to keep the obvious crap out of journals. It's an opportunity for paper authors to face objective questions from experts about things like their methodology, why did they do it this way instead of that way, are they aware of this potentially related work, why didn't they address this potential objection, etc.

It's also not unusual for scientific work originating in businesses to seek publication long after it has found application in the business. For example Google staff have published some important papers in computer science; but usually long after Google the product started using those techniques.

Now that they are under attack, Theranos' position would be greatly strengthened by having some peer-reviewed papers to wave around to show some legitimacy. But--the fact that they don't have those papers is not, by itself, proof that they are frauds.

It does seem likely that they are frauds--but that's because of direct investigation and reporting of their methods, not because of peer-review (or lack thereof).

Heh, this is, bluntly... untrue. In medicine, certain types of published studies are sometimes peer-reviewed, and sometimes people care about that, and that's about the extent of peer review as far as validity goes. Drug studies are not peer-reviewed in any meaningful way, and that's about as statistically strong a claim as exists in most of medicine. Medtech (implants, mechanisms, Theranos-like processes, etc) are certainly never "peer-reviewed".