I think one counter-argument to that trick was that it could be just as well interpreted as leveling by 1 each level instead of 2. And in case of uncertainty Caelic's holy commandments take the fore and since it's not Practical Op at all...
I don't see it as not being practical optimization. It's another way to create a double-threat if you throw on Ur-Priest at the end. There are certainly other uses to it. It's powerful, definitely, but it's not single-handedly going to break a game even if it will create a very powerful character who gets his spells earlier. Dragonwrought Kobolds can already get 9th level spells at levle 15 without this trick though, so moving it up to 10 isn't the end of the world.
At each level except 1st, 4th, and 7th, you gain new spells per day and an increase in caster level (and spells known, if applicable) as if you had also gained a level in both a prepared arcane casting class and a spontaneous arcane casting class to which you belonged before adding the prestige class level.
So, let's take Wizard with Spontaneous Summoning: I level up and gain a level "in both a prepared arcane casting class and a spontaneous arcane casting class." Let's apply the level I gained in a prepared arcane casting class: my wizard level increases by one. Let's apply the level I gained in a spontaneous arcane casting class: my wizard level increases by one.
That's definitely 2 levels, not 1. If it said you gained a level "in any prepared
or spontaneous casting classes" or something to that effect, you would only gain 1 level, but then again it would be ridiculous and be capable of advancing every arcane class simultaneously, allowing you to take Wizard, X, and Y and advance them all significantly (especially once you use Practiced Spellcaster trickiness to make them all have the lowest caster level by having the same caster level).
NOTE: Augmented Casting is non-functional with this trick since you only have one arcane casting class and it requires "another" class.
NOTE: Expanded Spell Knowledge is non-functional with this trick since your spell lists for your arcane and spontaneous casting class is the same list and it only affects the spell list, not which spells you can actually cast spontaneously.
So, while you definitely get a nice boost to spell-casting, two class features are entirely worthless, leaving two bonus feats, +4 boosts to caster level, and accelerated spell gain. It's still nice, but I'm still not seeing the "broken" that would push it into the realm of theoretical optimization
only.