It is exceedingly difficult to convince someone of a lie, if their livelihood depends on believing that lie.

A case study in human psychology for sure.

$$$ and NDAs work

I can't speak to Theranos, but having worked in scientific R&D the researchers are used to business folk telling them work is good enough/sellable even when they think it needs more study, are worried about edge cases, etc. And it can be true and not fraudulent. A lot of scientists are always looking for 'better', a more complete model, an answer closer to 'truth'; business is looking for what can I sell, what is slightly better (or differentiated enough) than current market offerings.

I think that can create enough plausibility in researchers' minds (along with keeping work divided and some psychological manipulation). They're thinking this looks kind of terrible and doesn't work for A, B, or C situations, but the business folks tell me this is exactly what they need and works on the market. Add in good pay (Theranos paid well above market for biotech for quite a while) and people won't think too hard about it.