Thread by @MerricB

In the early days of Dungeons & Dragons, the XP tables were designed so that characters went from 1st to 9th level relatively quickly, and then slowed down.

Gygax suggested that it take 40-60 sessions (a year of play) to reach level 9, then a gain of 2-3 levels a year after that

Meanwhile, Hit Point acquisition slowed down significantly after "name" level. Magic-Users gained a solitary hit point per level thereafter!

However, the power of magic-users (number of spells, etc.) kept increasing significantly.

I think it's fair to say that the various designers of D&D during the early days weren't sure WHAT to do with the higher levels. There was a structure there, but most of the game was aimed at levels 1-9.

The first time that someone really sat down and looked at the progression of D&D is likely for the BECMI books edited by Frank Mentzer.

I don't think 2E did that much at looking at the structure (though there's a little fiddling with it, it is pretty beholden to 1E).

3E took a much better look at it. Though it squibbed on the mathematics, so characters diverged too much.

4E is probably - to my mind - the edition that most looked at the progression from level 1 to 30. Though even then, the options for levels 21-30 (Epic tier) aren't as exciting as those for levels 11-20 (Paragon tier). Paragon tier is a masterpiece of design.

5E doesn't actually do a bad job of the higher levels (it's playable and fun), but it isn't a great job. The limitation on high level spell slots is significant, but available spells vs. monsters still cause problems.

The big shift of 3E is making it that characters progress from level 1-20(+) in a mostly linear fashion, rather than the sudden slow-down of AD&D 1E and 2E. That's something later editions kept.

I haven't really looked at the Mentzer Companion rules enough to determine how fast advancement should be - though I seem to remember the advice on XP does make it a faster progression than AD&D. So, in some ways, 3E may inherit the progression from Companion.