But an enemy being exactly at or below 7 or 8 hp is a pretty small range, even at low levels. I'm not sure I'm sold on the suggestion that you have an enemy precisely at that point in one in every three rounds. You've claimed that is your experience, and that's fair enough; I'm nonetheless not ready to call that typical.
I didn't make that claim. I said that the spell is useful once every three rounds at low level, especially for a Wizard that focuses on other area effect powers, including other At Will area effect powers.
Armchair aside, I've certainly played and run in tons of games - even without Magic Missile being present, it is usually noticeable when an enemy hovers in single digit hp, and it certainly hasn't been as common in my games as it has been in yours.
I think there will be times when that is the case, sure. But there will be plenty of other times when an enemy is at 11 hp, and another At-Will would be the better option.
Single digit hit points are not the only issue. You are missing the big picture here and focusing on that. You are not focusing on foes that have other number of hit points. Even a fresh first level foe with 25 hit points can often (nearly 40% of the time) be taken out quicker by Magic Missile (assume no striker attacks for this example), attack #1 hitting, and attack #2 hitting; then with Phantom Bolt missing, attack #1 hitting, attack #2 hitting, and attack #3 hitting. The odds of Phantom Bolt hitting and attack #1 taking the foe out are extremely tiny when the foe has 25 hit points, so attack #2 hitting is almost always required. The difference is that if attack #1 and #2 hits along with Magic Missile, attack #3 is typically not required. That saves a Standard Action to be used on a different foe at least one attack in three for this scenario. Sure, the DPR is similar, but there are initial hit point ranges where MM works better, just like there are hit points ranges where the extra few points of Phantom Bolt works better.
Let's take your example. Let's take a first level Wizard with Magic Missile (7 points at level one) vs. a first level Wizard with a D8+5 single target At Will power, and both have a first level Cleric with a D8+4 single target At Will power as backup. Note: there are D6 single target At Will attacks for both Clerics and Wizards, but there are also feats that can add +1 damage, so I consider this a wash for this example, not all PCs take damage feats at level one, not all Controller or Leader PCs do D8 damage at level one.
The foe is bloodied and has 11 hit points.
Wizard #1 with Magic Missile hits him for 7. He has 4 hit points left. The Cleric has a 55% chance to hit and kill, so there is:
0%: 1 attack and dead
55%: 2 attacks and dead
45%: 2 attacks and seriously wounded
0%: 2 attacks and not additionally wounded at all
Wizard #2 hits 55% of the time and criticals 5% of the time (60% chance to hit total). The Wizard has to roll a 6 or higher on the damage die to kill the foe. So, there is a 3 in 8 chance * 55% chance + 5% chance of the foe being killed outright in 1 round. The Cleric has a 55% chance to kill if the foe took damage from the Wizard and a 0% chance to kill if the foe did not.
25.6%: 1 attack and dead
18.9%: 2 attacks and dead
37.5%: 2 attacks and seriously wounded
18.0%: 2 attacks and not additionally wounded at all (i.e. requires 1 or more additional attacks from the Wizard and possibly the Cleric to take this foe down)
This is a hit point scenario that favors your POV (because the Wizard with a D8+5 power can kill the foe with a single attack) and there is still only a 1 in 4 chance that he will do that.
The foe is dead in a single round 43.5% of the time (with 2 PCs attacking it) vs. the 55% of the time with Magic Missile.
The foe isn't even additionally wounded by the two attacks almost one round in five. The foe is always seriously wounded in the MM case.
And what if the foe has 14 hit points instead of 11? The D8+5 attack is a lot less useful then because it cannot kill the foe with a single attack. In that scenario, MM wins more because the Cleric attack will almost always kill if it hits.
And, it's even more in favor of Magic Missile for Minions and for tougher foes (higher level or elites) that are harder to hit.
The reason it works this way is that the ratio of damage vs. the ratio of to hit. As an example, the D8+5 damage does 9.8 average damage (including criticals) vs. 7 for MM, whereas the ratio to hit is 6 to 10. 7/9.8 > 6/10. Until the 9.8 increases above 12 (2+ more damage per hit), MM wins the DPR race and hence, wins the "how often does it help more than a D8+5 power" race.
MM typically has a better DPR at low level and hence, there are a lot of scenarios (most of the non-area effect ones) where it is better. Even when the player boosts other At Will powers above the DPR of Magic Missile with feats and items, there are still a lot of scenarios where the Wizard either automatically takes out a foe, or sets the foe up for another PC to kill.
For example, the DPR of the other attack is higher than MM for a standard foe, but it is lower for an elite harder to hit foe.
But there is also the hypothetical scenario where your MM then leaves him with a few hp left, and was also a wasted action.
Except that it is NOT a wasted action. Here is where you are missing the point and mistaken. It actually does a LOT of good.
The Cleric with his D8+4 attack will automatically kill the 4 hp remaining foe if he hits.
The Thief or Rogue or Sorcerer can go after bigger game with their 15 to 20 point damage attacks.
Sure, one can play dumb and ignore the information and utility that MM gives the team when used appropriately. Or, one can play smart and use MM as a scalpel to either kill foes automatically, or set the foes up for a high chance of death by weaker allies like leaders.
And, honestly? I'd rarely expect to see a wizard breaking out MM on round 1 - if you've got one injured enemy and several other foes, now is probably when you should be breaking out your multi-target powers.
I agree that it won't happen a lot in round one, but it does happen. If there are other foes that an area effect can target, an area effect is better.
But, if there is a heavily wounded foe in round one (and at low level, a foe hit by 2 PCs or hit heavily by a Striker PC is often heavily wounded, often 75% damaged or more), then it can sometime make sense. The intent of other PCs is to throw out Encounter powers and focus fire in round one. Encounter powers often do more damage. The odds of there not being multiple foes in a close enough area combined with the odds of a single foe being heavily damaged is actually not insignificant. This can easily happen 1 encounter in 4 or better, depending on whether the DM plays the NPCs smart (i.e. avoid being in area effects).
My claim isn't that Magic Missile is bad, remember. My disagreement was with your claim that MM is so good that a wizard who isn't using it 1 out of every 3 rounds is playing badly. And I really haven't seen any examples thus far to back up that claim.
Again, Magic Missile is especially useful in the right scenarios. Just... not the ones you've really mentioned. For less optimized characters, it is on par with most other At-Wills. It is a useful tool to have on hand in the right situations.
But something that every good wizard should be casting 1/3 of the time? Not even remotely.
Every good Wizard who has it should cast it that often at low level.
And, Mages get it automatically. It really is the best choice for when it appears that it will kill a foe or when it leads to the death of a foe by a fellow PC when a different single target At Will Wizard power would not have killed the NPC. The rest of the time, it may or may not be the best choice.