The old complaint I'd always heard about the .30-06 AI was that it just used more powder to get to the same performance as the .30-06. Since the sides of the .30-'06 case don't have a lot of taper to begin with, the Ackley change only increases the case capacity by about 3 grains of water. My measuring records show a 203 grain .30-06 case having about 69.2 grains of water capacity and 203 grain .30-06 AI having about 72.5 grains of water capacity. This is not in full agreement with QuickLOAD's default numbers, which show even less difference.
At the time Ackley was making his rounds (pun intended), powder selection was more limited than today and the state of electronics was such that most folks didn't have access to a chronograph to measure performance difference. So the idea that the performance was the same would have most often been based on trajectory and observed results on game, which must have appeared to be about the same.
I don't own one of the AI incarnations of the '06, so I worked up equal pressure loads of the same powder and bullet in the '06 and '06 AI in QuickLOAD, using the case capacities I mentioned earlier. The velocity predictions only give about a 35-50 fps advantage for the improved version of the '06. This is regardless of whether its a 150 grain bullet over IMR4895 or a 180 grain bullet over IMR4831. That difference is within muzzle velocity extreme spread for many loads, so, assuming it is an accurate comparison, you can see why folks without chronographs might conclude there was no advantage to the AI version?
Had Ackley chosen to shorten the neck to the length of a .308 neck, then he'd have got a bit more room and bang for his buck with the the '06 AI. But neck shortening isn't part of his formula for improving cases. Besides, back when the round was devised, most folks took it as gospel that a shorter neck meant poorer bullet alignment and was undesirable. Ackley devised the .30-06 AI in 1944, but it wasn't until Middleton Thompkins win of the National Highpower Championship with a .308 in 1963, repeated in '64 and again a few non-consecutive years after that, did this bit of ballistic dogma begin to be questioned. I was still hearing it about the .223 and .222 in the 1980's, so it takes awhile for some of these ideas to be put to rest.
If you want a .30-06 AI, I think your purposes have to be the AI form's reputation for better case life and lower bolt thrust. Figure to budget an extra 4% for powder.