Hardware is hard. Let's go shopping. :)
However, that doesn't have to be the end.
You can narrow the usage to something like "Use for routine visits to collect data quickly but not for diagnostic reasons". Being able to do a very fast test without drawing much blood for every time you visit the doctor would be useful information over time. There's also a "Hey, that number is a bit high. Let's pull a real amount of blood and see what's going on." I suspect that if you correlate the information over multiple tests it gets much more accurate very quickly.
In addition, I suspect that the engineers would have eventually worked out how to correlate things correctly and figured out how much blood was required for what level of certainty.
Finally, most people in the biotech sector understand that it's a "pump and dump" scam. You invest, wait for the company to get some good news, and sell to the next layer. Biotech often takes long timescales and different investors toss money in at different points.
However, what doomed the company was Elizabeth Holmes lying. I could make other comments, but, really, Theranos would still be fine if she had told the truth. People know that biotech always runs into problems. Always.
Your CEO is one of the people who must not lie (your CFO is the other). Your sales people can lie. Your marketing people can lie. Oddly, even your CTO can get away with some things, but normally the CTO is a shining beacon of truth compared to most people in a company. If your CEO always tells the truth, the CEO can say: "We've hit some snags, but the engineers think they can work around them" and people will believe because they want to believe.
But once your CEO is known as a liar, you are doomed.
However, I expect that Elizabeth Holmes will be back far too soon and will wreak havoc on an even larger scale. She is just Carly Fiorina writ 35 years later. She will fail upward because she is a not-unattractive blonde female with lots of rich family connections.