We have had this discussion a few times in the past. The last time I actually did the math on example magi from prior editions as well. Outside of 4th edition, the average was always a touch above +2. 4th was the outlier specifically because its rules made Intelligence much more important. There are many blocks of text in 5th which are identical or with minimal changes since at least 2nd edition (don’t have a copy of 1st).
Published numbers often do not make much sense when actually looked at and worked out. It is not just like “Average Magi Int is +3”. You also have things like the number of gifted individuals which with the default rate should only be about 7,350 total, one for every ~535 square miles. That is total, which includes not just Magi and hedge traditions but all the undiscovered.
Somethings actually produce useful material when the numbers are worked out, like the Aura rules in RoP:M which gives you good rough estimates for how magical auras and regio grow and change over time.
What it all comes down to is do you want to just use the published numbers, even though they might not make sense if you run the math. Or do you actually want to run the math, which requires you to ether change the world to make the math make sense or change the math to fit the world. I have done a lot of the latter.