If you chose a high estimate of feat = 2nd level spells, the fighter line's height is obviously doubled. This puts fighters' progression just below the wizard one at level 8. Does anyone feel the need for me to recreate such a picture? I can imagine it very easily. The shape doesn't change, just the f(x) = g(x) point.
That covers part of my original point, but not all of it. The first feat in a typical feat chain is about equivalent to a second level spell, yes: however, the higher feats in the feat chain are worth more. Even in a low-optimization environment, fighters tend to pick feats higher up in the same feat chain rather than the start of another feat chain, if possible. So the first feat a fighter takes would be equivalent to a 2nd level spell, but the second and third would be higher. This continues until he completes the feat chain and must start another one. Thus the shape would be an upward-rising curve, followed by a plateau and then another upward-rising curve.
The issue and what PBMC is attempting to illustrate is a long-known fact about casters, which is that their power growth via spells exclusively is entirely exponential, and is further pocked by spikes above the curve at each level - your Polymorphs, Celerities, Time Stops, Magic Jars, etc. While a feat chain can certainly ramp up in power with each further feat, they do reach an inevitable plateau and cannot advance further, meaning another feat chain - unless directly synergizing an existing feat chain - would be a lateral move and its own ramp up and plateau. Spells on the other hand, are as a single mechanic their own exponential curve, and while there are certainly plateau points at each new level of spells, there are equally selections within any given level that trivialize anything the caster may encounter at level-appropriate CR
and fall within normal mean optimization. Even a John Q Public Wizard will likely have Sleep, Stinking Cloud, and Black Tentacles by level 9 or 10, as a few examples, and that's only a VERY small example of the bigger whole from the PHB only. Another example of a spike above the curve is Grease - a 1st level spell that hoses chargers (anything ground-bound for that matter) and makes anything involving climbing a death sentence. And that's just ONE SPELL that most Wizards, even the fire-based stereotypical blasters pick up.
Part of the complexity in quantifying spell power has to do with the big 3 of the 4 common Wizard roles being mostly force multipliers, however as Sor0 demonstrated, even a rudimentary buff sequence that should be relatively common can turn a caster into a melee powerhouse with minimal effort. Even without feats augmenting their capability, they are able to fulfill the ROLE needed at that time, while still having on tap debuff/blast/BFC capability.
As for fulfilling multiple roles at once, it's relatively easy to debuff, blast, and BFC with a SINGLE SPELL with even a modicum of thinking on the part of the Wizard player. The closest thing to that a Fighter has is a chain tripper, most likely, and while they can lock down like a champ, they still actually RISK something to do it. In particular once a Wizard can do the above from greater invisibility or with a Mirror Image running, the odds of there being reprisal for his action are laughable.
And this also isn't taking into account the force multipliers that are PrCs as well, which takes all of the above to a whole new ballgame.
@Faeryn : I think describing it as cubic or exponential with quadratic options (the spikes above the curve) might be slightly more accurate, but the end-result is roughly the same. The leap attacking demoralizing chain tripper reaches an absolute plateau, while the wizard's options keep getting increasingly more powerful than the level before, up to the endpoint of access to the abilities of any creature ever, being able to stop time outright, and being able to call in beings far more powerful even than themselves.
