How credible are the reports that Landis' results included synthetic test? The linked espn story attributes this fact to the head of the French anti-doping council. I don't know enough about cycling to know whether this is the relevant authority, or whether the synthetic test results have been verified. But this seems to be the crucial issue.

From what I understand, the doping agency tests the T/Epi ratio and his ratio fell outside the 4:1 acceptable range. Fine. One of the rumored justifications for the test is the alcohol consumption and cortizone injection (for the hip) the day before the positive test, both which lower Epi and would therefore increase the ratio without any increased levels of Test. Obviously, if the reports of synthetic test prove to be accurate, it is tough not to accept the result. But until then, I'm going to stay cynical ...

I see now that Landis (on Good Morning America) has distanced himself from the whiskey story (among others), and seems to acknowledge that the test does indeed include unnatural test. The defense is now apparently exactly the strange observation that Gordon points out: that only one of 8 tests came back positive with negative results both before and after the stage in question.