They lowered the threshold so it should catch more people at risk for diabetes. I think they should use a1c as a diagnostic tool. It doesn't mean they should only use a1c to test people. Just like they shouldn't only use fasting glucose to see if people are fine, which is what is currently going on right now. But we all know that for some people, impaired glucose tolerance happens way before there is an increase in fasting glucose. And an a1c will help to catch these people who may be unwilling to sit two hours for a OGTT.
So yeah, person #1 with prediabetes with a fasting glucose of 100 can have a normal a1c of 5.5.
But person #2 with a normal fasting glucose of 79 should generally not have an a1c of 5.8 unless he/she is at risk. Using a1c as a diagnostic tool will not catch person #1 but it will catch person #2 which is was the focus should be on. Person #1 was already caught with the fasting glucose which is what doctors in general, always order.
Now, a great doctor will order an OGTT. But I imagine patients aren't very compliant with an order of having to sit at a clinic for 2 hours, getting blood drawn thrice. A1c is the next best thing. At least, it'll give the doctor some idea of likelihood.