In short, 2 unoptimized Sorcerers head to head, Vancian wins.
Optimized, Dynamic caster will win.
Dynamic casters 'Spell Pool' is their nonlethal damage. Since they only take nonlethal damage on a failure to save (Special Fort save vs spell casting DC)
Whenever a character casts a spell, the character must make their Fortitude Save to resist Spell Drain. The DC for the Fortitude Save is given in the spell description of the spell being cast. If the characters passes the Fortitude Save, they are unaffected by the effort of spellcasting. If they fail, they consult the Spell Drain Table to the right, and based on the DC of the spell being cast, suffer the non-lethal damage shown.
Emphasis mine. There is an earlier segment that refers to all attempts requiring Spell Drain, but it would fall under the Specific trumps General clause, as the system is also compatible with other d20 systems that possess an energy stat that can be used for spellcasting purposes. Since 3.5e does not have this stat, it is covered with the specific nonlethal damage guidelines.
Given that the Save DCs range from 15 to 51, with additional modifiers for spells with costly component, xp, or time costs(greater than 1 round), It introduces a fairly high failure chance, even with the caster receiving a bonus to this save similar to base attack bonus progression (with primary casters receiving the highest rate). i.e. Sorceror receives a +1 to Fort save DC per level when applied to casting a spell.
The Spellcasting bonus is a bonus that applies to Fortitude Saves, Drain checks, and Control checks.
So an average unoptimized level 10 caster has a fort save of +3-5, 13-15 for spellcasting fort saves. That gives them a 50% chance of failure on a level 2 spell (Casting DC 25). 5% Chance of success on a level 5 spell(Since classified as a fortitude save, it is subject to critical success/failure).
Furthermore, casting DCs are just as difficult, since it is only 1D20 + spellcasting bonus + relevent ability modifier, compared to the same casting DC.
Level DC Failure(avg) >10 Failure
0 15 1d4(2.5) 2d4(5)
1 20 1d6(3.5) 2d6(7)
2 25 1d8(4.5) 2d8(9)
3 30 1d10(5.5) 2d10(11)
4 34 1d10(5.5) 2d10(11)
5 38 2d6(7) 4d6(14)
Since an average level 10 sorcerer only has 26 hit points(assuming avg con), they cannot handle very many failures before unconsciousness.
If the Fortitude Save is failed, a Control check will be required. If the Save is failed by 10 or more, the spellcaster becomes fatigued. If the character was already fatigued, he or she becomes exhausted instead. If the character was already exhausted, he or she falls unconscious.
A spellcaster passing out from spellcasting remains unconscious for 1 round per Point of Drain caused by the last spell that was cast. That is, an exhausted character who fails his or her Fortitude Save by 10 or more and suffers 30 points of Drain will remain unconscious for 30 rounds ? 3 minutes. After waking, the character will still be exhausted.
Side By Side - Without Optimization
Casting Ability Single Combat (No bonus spells for stat or specializations, assume average rolls and damage for Dynamic). Assume average con, average HP of 26.
Level Vancian Dynamic
0 6 104(5%fail, 2.5 dmg,0%dmgx2)
1 6 22(35% fail, 3.5dmg, 0%dmgx2)
2 6 8(60%failure, 4.5dmg, 10%dmgx2)
3 6 4(80%failure, 5.5dmg, 30%dmgx2)
4 5 3(95%failure, 5.5dmg, 50%dmgx2)
5 3 2(95%failure, 7dmg, 70%dmgx2)
*Dynamic caster only needs 3 rolls that fail by more than 10 to lose conciousness(of which 2nd level spells have a 10% chance)
Side By Side - Optimized
Level Vancian Dynamic
0 8 528 (5%fail, 2.5 dmg,0%dmgx2)
1 7 378(5% fail, 3.5dmg, 0%dmgx2)
2 7 293(5%failure, 4.5dmg, 0%dmgx2)
3 7 240(5%failure, 5.5dmg, 0%dmgx2)
4 6 240(5%failure, 5.5dmg, 0%dmgx2)
5 4 62(15%failure, 7dmg, 0%dmgx2)
With the Dynamic caster optimized, they have about 2.5 times the HP pool of the Vancian caster, since CON shares top priority for them(CHA/CON priority), whereas con usually ranks lower for Vancian.
With adequate optimization(including magic items), a Dynamic caster can outcast a Vancian sorceror. In addition he will have more HP, and better saves. He will definately be far weaker than the Vancian in the early levels (prior to being able to afford the range of good save bonus gear), though the difference will become far more pronounced as he levels.
However, over the long haul, the Dynamic caster will still outperform the Vancian caster due to natural nonlethal recovery (1HP/LVL per hour).
Daily Spell Usage
Level Un-optimized Optimized
0 1064 2448
1 218 1749
2 74 1360
3 32 1113
4 23 1113
5 14 292
This system likely works really well with other systems that do possess an 'energy' stat, but it needs far more in depth balancing to end with comparable power to the Vancian type. I also only spent about an hour identifying optimization options for dynamic spellcasters(only got fort save to 25 before spellcasting bonus), there are likely dozens more to further widen the gap(without munchkin/cheese). The rest was math.
Now, thanks to saving throw guaranteed crit/fail rules, the Dynamic sorceror will always have a 5% failure chance for any spell they cast, at which point they get a second chance successfully cast the spell with full consequence...using an attribute that cannot be truly optimized. A level 20 sorcerer with 26 charisma has Casting check of d20+20+8, DC51 is practically impossible, especially since by RAW checks generally can't critical.
Bonus - Per RAW, wish could be cast without an XP cost...assuming the sorcerer could make a DC 101 fort save. Some quick math shows a 10% success rate with the majority(752k) of his wealth in save gear.