Real life experience:
Fasting BG 138 and 3 months later, 131. Diagnosable as diabetes? Yes, currently, but then, no.
A1c 2 months later, 4.8. Diagnosable as diabetes? NO!
If the doctor had relied on the A1c to treat me, I would not have received appropriate treatment. My doctor was satisfied with fasting BGs in the 150-180 range and did not consider postprandial rises, because they came down in time for the next meal. And my A1c was ALWAYS in the 5's and 6's, even as my BG readings crept up and up.
We tried sulfonylureas, which didn't work, so I asked to be put on insulin. After a lot of trial and error, including 70/30, which resulted in 2 hypos per day unless I was constantly snacking, mixing NPH and R, then NPH and Humalog, and finally, I asked for a pump. He refused the first time I asked, but acquiesced a year later.
I have been very happy on the pump.
A year ago, my BGs started to rise and wouldn't come down even with large doses. 3 months ago, I had an A1c of 10.1 and a month later, 10.7. I ended up in a coma a week after I saw my doctor with the 10.7 -- he never twigged to the fact that it was a life-threateningly high A1c for me. If not for friends missing me at a picnic and coming to check up on me, I would have died.
I know other people have had A1cs higher than that with no serious immediate consequences, and I know what the figures are for A1cs and BG equivalents. But the problem is that I'm a low glycator, and my A1cs indicate my fasting BGs, not average BGs.
It doesn't matter what the statistics say or what expert committees believe, in the treatment of diabetes, sometimes the hoofbeats you hear ARE zebras!