Well, the metrics here are base tier power(played naive, sticking to the fluff suggestions), maximum power(which runs up against the issue of character vs class derived power and how much you could probably get away with in an actual game) and effort to gain ratio. Looks pretty hard to quantify into a neat numerical ranking. Theres also ease of actual play.
The genius part is things don't change all that much.
For instance you rank the Cleric T3XYZ because it heals. Well, when? During combat wouldn't even validate that T3 mention once they have access to Heal which does out scale a monster's damage. Negating the entire attack sequence of a monster is virtually better than trying to use a debuff to the same effect since they could bring their SR or Saves into account. It's a far cry from the best option, but Clerics also naturally obtain Save or Die spells they can utilize during the first round, as well as summon spells and even dangerous Mark spells which can be precast on items and tossed (seen here
) or even carried in hand and dropped as a free action. And just what is it they do out of combat besides solve you're problems? Well conically, they are the only
one who heals or raises dead other than the depreciated Healer. Pretty unreplaceable I'd say. So pretty much end back up exactly where they started, more useful than most other classes.
Take the Fighter, Barbarian, Monk, all they do is fight. Everyone
has combat talent, even domesticated animals, D&D is a combat oriented game. They struggle to even preform their role too, ever seen a Barbarian forgo charging? Did he come close to the power of Strikes? How about non-attack/damage effects? Truth is, they still suck and have to optimize to keep up with the Druid's spellbuffed pet.
And things don't change all that much in maximum power either. Which one is "stronger", Wizard or Sorcerer? More spells vs knows them all and learns the a level sooner. No real contest. You'll find placement is again the same. A few nitpicks on certain classes could be found if you really got into the details, under any generalized measurement, but the average comes out to the same thing. If anything, the additional details prompt people to argue those points opposed to a more generalized description.