And I thought the phrase "too big to fail" referred only to large corporations...
Yes, but now that the detection limits for CB are being pushed to ever lower levels, it might be possible to extend that window further. Drugs like CB concentrate in the hair follicle, from which they gradually leak into the growing hair. So when the hair is cut, all the drug does not disappear; there is still a residual amount in the follicle, which will continue to leak into the newly sprouting hair. In theory, with a low enough detection limit, one could detect a drug in hair over a period of many months and several haircuts. Thus IN THEORY it is POSSIBLE that EVEN NOW Bert's hair could have some CB in it. Or that he could have transfused in early 2010 and still tested hair positive in August or September.
In actual practice, I don't know (and I very much doubt anyone else does, though it probably could be calculated) how long the window of detection in hair could be. My guess is you are probably right, and that when Bert was notified, it would have been unlikely that CB even from a hypothetical post-Dauphine transfusion in June would have shown up. Certainly that unlikeliness is a good enough rationalization for not submitting to the test.
Yet as others have pointed out, submitting to the test would have been good for his PR campaign. Looking at those documents his lawyers submitted, it was clear they were using any argument, no matter how scientifically tenuous, to proclaim his innocence. A negative hair test would have fit in very well there.