She would permit outside researchers at the esteemed Cleveland Clinic to study her technology—zealously protected trade secrets—and publish their results in a peer-reviewed journal, she announced at a conference at the clinic on October 27.

But on Monday night, nearly 100 days later, Theranos spokesperson Brooke Buchanan acknowledged to Fortune, “We haven’t started yet.”

The company has no timetable for beginning either, Buchanan admits.

“It takes time to get these studies in place,” she explains. “All the right decisions. We should know in the coming months.”

The lack of progress was first reported last Thursday, in the ninth paragraph of a Bloomberg View article about Theranos’s reportedly fraying relationship with a key commercial partner, Walgreens. There a Cleveland Clinic spokesperson told reporter Caroline Chen, the clinic “has not begun any work related to the study at this time.”

Holmes’s decision to permit the study deflected and defused pressure that was building after the first of a series of Wall Street Journal articles, published on October 15, cast doubt on the accuracy of Theranos’s tests, and even on the frequency with which the company used them—rather than simply employing conventional needles-in-the-arm and analyzers purchased from third-party manufacturers like Siemens.

“Data is a powerful thing because it speaks for itself,” Holmes said at that time, as reported in a New York Times story entitled, “Theranos Chief Yields to Calls for Proof of Blood Test’s Reliability.”

Meanwhile, in December, in what looked like a second nod toward greater transparency, the company also seemed to pledge to take a second step. A Bloomberg Businessweek cover story that month, for which Holmes provided extensive cooperation, reported: “Theranos is preparing ‘manuscripts’ containing the testing data that’s been submitted to the FDA, which it plans to publish in a medical journal.”

But Buchanan tells Fortune that there is no plan to do that—at least not unless and until a given test has already been approved by the FDA. (The company says it has filed preliminary submissions to seek FDA approval for more than 120 of its tests.) “Once we get them back they’ll be published, but not before,” she says. “We never said we were going to do it before.”

She said the company is, instead, permitting interested journalists to view the confidential data, and to be “walked through it” by Theranos personnel, though it will not provide physical or digital printouts of the data itself. (Theranos showed me scores of pages of validation data in Spring 2014 but, as a journalist, I lacked the training to assess its significance.)

In that sense, she says, the company is publishing the data. “I don’t know what you mean by the word ‘publish,'” she says. “We’re making it public.” [See also addendum at end.]

In fairness, Theranos has more urgent priorities at the moment. On Thursday it will file a confidential submission with the Centers for Medicare and Medicaid Services (CMS) attempting to assure the regulator that it has adequately addressed five major deficiencies the CMS found at the company’s Newark, Calif. lab during an inspection last November and December, one of which it judged so serious as to “pose immediate jeopardy to patient health and safety”—i.e., “a situation in which [the deficiency] has already caused, is causing, or is likely to cause, serious injury or harm, or death.”

If CMS is not satisfied with Theranos’s response, it has the power to impose fines, require on-site monitoring, or suspend or revoke the lab’s certification to operate. (In a statement issued when the CMS action came to light, Theranos asserted that it had already addressed many of the deficiencies, and stressed that the problems did not affect a second lab it operates in Scottsdale, Ariz.)

The prospect of having the Cleveland Clinic authoritatively scrutinize Theranos’s technology has understandably attracted considerable excitement within the health care community and the press.

Dr. Delos (“Toby”) Cosgrove, the CEO of the Cleveland Clinic, spoke about the planned study with Theranos on CNBC’s Squawk Box on October 30.

“I think what we need to do is have verification of the technology,” he said. “We don’t have that, and we have made an agreement that we would begin to test that with her and publish the data. . . . Somebody has to verify this.”

(Back in March 2015, Theranos and the Cleveland Clinic announced a “long-term strategic alliance.” The clinic’s researchers can nevertheless be considered independent in that “no money exchanged hands” in the alliance deal, Cosgrove told CNBC, and the clinic has made no investment in the company, a spokesperson told Fortune.)

Buchanan told Fortune Monday night that its relationship with the Cleveland Clinic remains “strong” and “multifold,” and that the two organizations meet “weekly” if not “daily” to discuss a range of subjects.

Asked to characterize its relationship with Theranos, Cleveland Clinic spokesperson Eileen Sheil said, “Originally we were going to validate their technology, but that has not happened yet.”

Addendum:

After this story went to press, Theranos spokesperson Buchanan called to stress that the company does plan to publish in a medical journal the results of a “roundtable” that the company is holding—also alluded to in the December Businessweek cover story—in which, according to the magazine, certain medical experts are being invited to visit the company for two days “to examine the company’s technology, data, and regulatory filings, and . . . then talk publicly about what they found.” The roundtable was originally scheduled for December, but has been put off to late February or early March because of a family issue relating to one of the experts, she said.