It’s a legacy of old-school D&D which was kept in ACKS. It definitely does reward the already strong and in that sense it’s counter to the modern trend of game design. For example, Mekton Zeta allows you to start with great characteristics OR to get a big xp bonus).
The original rule does have a few subtle benefits, though.
First, in a game designed for 4-8 players, niche protect is an important design goal. D&D-type games that match class and characteristic systems create distinct niches for the PCs. The strong fighter doesn’t have to worry that the mage is better at smashing things than him. The dextrous thief doesn’t have to worry that the cleric is better at missile weapons.
The XP bonus for prime requisites helps encourage PCs to select a class that aligns with their ability scores, and thus encourages niches. PCs will tend to want to raise their prime requisite to reach the XP bonus if they can. Since raising their prime requisite by 1 point requires lowering another ability score by 2 points, this will tend to “flatten” their non-prime scores.
For instance, consider Albus (Str 14, Int 15, Wis 9, Dex 10, Con 13, Cha 8) and Bradwell (Str 7, Int 13, Wis 10, Dex 11, Con 12, Cha 12). Because of his Con, Albus decides to be a fighter even though he could also be a mage. Bradwell has little option but to be a mage even though he’s not as smart as Albus!
Albus decides to lower his Int from 15 to 11 in order to bump his Str from 14 to 16. Maybe he’d have done that anyway, but the fact that it means he’ll get a +10% bonus to XP certainly helps make it a smart decision.
Meanwhile Bradwell lowers Dex from 11 to 9, Con from 12 to 10, and Cha from 12 to 10 to get his Int from 13 to 16. Perhaps Bradwell would have done this anyway, but again, the extra bonus to XP makes it even a smarter decision.
And thus Albus ends up less intelligent than Bradwell (Int 11 v. Int 16) even though when the dice were rolled Albus was more intelligent. Niche protection! The mage doesn’t feel like he’s dumber than the fighter.
A second benefit of the XP bonus from prime requisites is that it tends to prevent the hybrid classes (such as Elven Spellsword) from advancing quickly or being as hyper-effective as more niche classes. For instance, if Albus had decided to be an Elven Spellsword, he could not have lowered Int to raise Str, or vice versa, because both are prime requisites of the Elven Spellsword. He could lower Con from 13 to 9 if desired, but even so he cannot get both stats to 16, and so he cannot get the +10% XP bonus. At best he can be Str 15, Int 16, or Str 16, Int 15, and since XP bonus is based on the lower of the two prime requisites, he’ll be at +5%.
Anyway, all of the above is VERY subtle, and I can imagine many campaigns where you’d want to drop the rule to create a different flavor. If you want your campaign to encourage oddball heroes and ordinary-men-who-rise-to-the-occasion you could implement the system in reverse. For instance, at 9-12 in your prime requisite you get +15% XP, at 13-15 +10%, 16-17 +5%, and 18 0%. The in-world justification would be that “the second best try harder”.
Just be aware that it will slightly tilt the balance towards hybrid classes and against niche protection.