My first charge, after complete discharge (so this is the "cal" cycle), took >6 hrs (not sure total, as I went to sleep before it was done).
I wasn't too worried about it, but reading some still had extremely long charges on subsequent cycles, I decided to do a quick test.
I ran the battery down to 50% of capacity (showed a bit more than 4800mAh), playing YouTube videos, and music, while screen sharing.
It ran [on battery] for 1:20 (to 50% capacity), so that's not too unexpected considering the type of use.
Plugged it in, and the recharge time settled in on 1:30, from about 50% full. This seems about right, to me.
edit:
Update -
After about an hr, I checked back (using screen sharing - MBA running during charging). Now it was showing 1:40 to full charge, from 83%. I checked the charge rate, and it was down to 0.5A. So, it looks like it goes to low rate sooner (% wise) than I've seen with the MBP. Just running the MBA takes that, or more, so it makes sense the total charge time will increase noticeably after switching to low rate (typical of any machine, though).
It looks like this will be normal for the MBA. Now, it's been sleeping for about 20 min. Woke it up to check remaining charge time and it reduced by 40min (about an hour left). If I leave it sleeping, it'll no doubt be finished much less than that.
I think what people are seeing are charge times running vs. sleeping/off. The 45W supply just can't keep up the higher rates and also provide the overhead to run the machine. All machines will charge faster sleeping or off, but the difference is going to be more apparent with the MBA. I suspect they want to keep the internal component power levels down (temps - maybe smaller gauge wiring, traces, and components themselves).
I haven't seen any posts from anyone trying the 65W supply, but I'm going to guess the times will be the same [as the 45W]. The charge system should limit current, so it doesn't matter if you have a bigger supply.