I'm not in favor of a pure % aip game, because the power of the champion is constant (in that it raises at a pretty standard rate) while aip fluctuates so much in games.
With aip floor riding game 20% is barely noticable, but in a game with exo waves or a fallen spire games 20% is lethal.
The relative impact of champions is also inversely proportion to aip. aip is often the result of the human player getting more powerful. To illustrate.
The power of the champion when the player is at 10 aip (one planet) is exponentially more powerful relatively then a player at 220 aip (10 player worlds + other aip games). But wait, you can reasonably say, champions get upgrades! This brings up the next point.
Further compounding the issue is that with various tactics the champion can increase in power without increasing aip. So map placement is very important. Some maps may with clever planet hopping allow the player in a 80 planet game to get cruisers with only 4 or 6 planets taken by the player. To balance this, the current model of having nebulas tied to planet count is done, but that results in players not getting the full AS experience. Yet the alternative of always having 8 nebulas regardless of planet count results in the player being able to get cruisers without taking a single world which is its own problems. I already can imagine games where players on a 10 planet game assaulting ai homeworlds with cruisers and wearing them down.
So with these points, I am not in favor of a pure % in aip due to the massive fluctuations of map types, number of planets, play styles resulting in so many varying ai responses. Fallen spire causes a drastic change to the game, but it causes the same responses on its campaign no matter what map type, number of planets, or play style, because its difficulty is not tied to aip but rather ai difficulty numbers.
