See my errors 3 and 4. Metamgic isn't for every spell. And it's not for the same spells at all levels. But your conclusions are seriously mistaken.
Higher level spells vs. metamagic spells:
At level 5: Vampiric Touch vs. Empowered Shocking Grasp: 2d6 (get temporary hit points)--avg damage: 7 vs. 5d6x1.5 (avg damage: 26). Advantage: Empowered Shocking Grasp.
At level 5: Ray of Exhaustion vs. Empowered Ray of Enfeeblement: 6 str, save for 2 vs. 1d6+2x1.5 (avg: -8 str, no save). Advantage: unclear since Ray of Exhaustion has important effects on dex and mobility... but they stack so why have them compete. At level 10, the empowered ray of enfeeblement averages about 12-13 points of strength and the save is easier for enemies to make so I think the advantage shifts to the empowered spell.
At level 7: Empowered Scorching Ray vs. Enervation: 1d4 negative levels vs. (effectively) 12d6 fire damage. Advantage (IMO) scorching ray
At level 7: Empowered scorching ray vs. energy orb spell (T&B or CA or mini HB): 7d6 plus minor effect (save to avoid) or 12d6. Advantage: Scorching Ray
At level 9: Empowered Fireball vs. Cone of Cold: (effectively) 13.5 d6 vs. 9d6 with a DC 2 points higher. Advantage: Empowered fireball...in almost all situations. (The best case for the cone of cold is probably a 10th level halfling rogue (+13 ref, and evasion which will negate the advantage of the extra damage on a successful save). Assuming an int of 20 and greater spell focus, the DCs for the fireball and cone of cold will be 20 and 22 respectively. Thus, the rogue will take an average of 14.175 from the fireball and 12.6 from the cone of cold. It only gets better for the empowered spell from there on out--against a foe like a 10th level cleric with a +4 reflex save, the empowered fireball does an average of 41.34 damage. The cone of cold does an average of 29.14 damage--a pretty dramatic difference). [My method for calculating average damage: odds of saving x damage on save + odds of failing the save x damage on a failed save]
At level 11: Maximized Scorching Ray vs. disintegrate: 72 points of damage vs. 77, save for 17.5. Advantage: Maximized scorching ray (and it's a level lower) unless you want to disintegrate a door or a wall of force.
And those are the spells that are easy to quantify. Other metamagic feats are harder because there are no directly comparable spells. Quicken Spell, for instance is a great feat IMO but doesn't have any really good comparison points because there are very few free action spells and none in the PHB. But, name me a fifth level spell that can be cast as a free action and deals 1d6+5 points of strength penalty to a single target with no save. Not as spiffy an effect as wall of force to be sure, but you can cast a quickened ray of enfeeblement during the same round as you cast a wall of force--a huge advantage. For an 11th level wizard using a quickened scorching ray, the ability to dish out an extra 42 points of damage (on average) for several rounds is very significant.
In short, metamagic spells worth casting are almost always better than normal spells of the same effective level (that's what makes them worth casting). They're either better because they do more of whatever they do (empower and maximize spell) or because there's no other way to do what they do (quicken spell).
reanjr said:
For Wizards, it is rarely cost-effective to metamagic a spell. It is almost always more advantageous to simply memorize a higher level spell. There are some specific cases that can be planned for that this is not the case, but they seem to be the exception rather than the rule.
For sorcerers, without bonus feats with which to take metamagic feats and with the increased casting time, it just doesn't seem like a popular option. In my opinion, sorcerers should get 5th, 10th, 15th, and 20th level bonus feats for metamagic and should not increase the casting time. They're already weak as it is.
Wizards need a way to use metamagic feats without using up those high level slots. The main problem I see is that the only spells that are low enough level to metamagic are also the spells that seem most off-kilter on the resultant level. For instance, casting a maximized ice storm (7th level) for 30 points of damage to anything within a 40 ft. cylinder seems a bit weak. But compared to a maximized burning hands (4th level) at 20 points of damage to a 15' cone, it's powerful. And are you really going to give up one of your high level slots? Probably not. You're much more willing to waste a 4th level slot on a weak spell than a 7th level one, if only it weren't that much weaker and useless. There are too few spells and situations that are worthwhile. This problem is exacerbated by stacking metamagics (a quickened maximized magic missile would be nice to have, but not as an 8th level spell).
There are several different things that could be used to replace the spell level adjustment of metamagic. Casting time, expensive and specific material components, XP, etc. I think any of these would be more viable options than increased spell level.
But there is still the buy-in cost of getting a metamagic feat, though for wizards, this shouldn't be much of a problem.
In summary, it's because metamagic feats suck. They're a neat idea, but horribly balanced.