I think that something like this could work- with or without Crothian's modification- depending upon the other rules and assumptions in the game.
A game with a 3.X-style emphasis on balance at each level would necessitate cooperative leveling.
But examine a game like RIFTS (hear me out!) which has a class/level structure like that of 1Ed D&D, in which game balance is not a primary concern. PCs can have radically different abilities at first level, but progress at rates commensurate with their power levels. In other words, a "weaker" PC would advance more quickly than most others, and would improve significantly with each level. Meanwhile a "front-loaded" PC would advance very slowly, and level-based improvements might be only marginal. "1st level" or "10th level," thus, means different things to different PCs. Cooperative leveling might not be neccessary.
A game in which PC design is based upon pure archetypes or a point basis, even "leveling up" might be virtually meaningless in the traditional sense. In a game like that (assuming point base) a 250 point Combat Monster archetype will perform radically differently from a 250 point Ultimate Detective archetype. The latter may not even be able to survive 3 seconds on a battlefield the former could sleep on- meanwhile the former would find the Mystery of the Open Bag of Cheetoes completely insoluable even while the latter is surreptitiously wiping his power-orange fingers on the former's cape. Each would improve in their own fields of expertise, but comparison between the two might be meaningless. Cooperative leveling might not be neccessary. In fact, it might be better for the game if PC improvement were story driven- you improve what you used- in a non-discrete, "analog" kind of PC advancement progression. The PC who only uses his combat skills only improves his combat skills, and does so quickly. The PC who uses everything on his sheet on a regular basis might have across the board incremental improvements.