The auto-hit magic missile was also extremely low availability compared to a sword. Given the AD&D 1 minute round, a fighter could be using his sword 200 or 300 times a day. That wizard can't.
This is all in-fiction. But we are not interested in the imaginary tension experienced by the imaginary characters in the imaginary gameworld. We are interested in the real tension experienced by the real people in the real world, sitting around a kitchen table playing the game.
For them, swords aren't particularly more prevalent than attack spells - as best I'm aware, wizards and fighters are the two most popular classes in the game, and have been for most of its history. And I wouldn't be surprised if wizards are more popular than great weapon fighters (though that is pure intuition, resting on the fact that GWF means forgoing the shield, which is one of the fighter's class specialties).
MM's tension isn't the miss, it's the decision of when to use it.
Similar situation for all AoE spells. The damage on a miss there (more correctly reduced damage on saving throw) was a way of making the decision less agonizing.
In 4e, the at-will magic missile was changed from requiring an attack roll to auto-damage, mostly (as best I can tell) because the customer base wanted it to be auto-damage as it had been for the previous few decades.
In both 3E and 4e, a wizard could use an auto-damage effect bascially every round of combat. Likewise in AD&D from around 5th level, by which point, between spells and wands, the MU would generally have enough spells to cast one every round.
The meaningufl decision point for the player of a magic-user isn't whether or not to deal damage, it's which spell to use (based on considerations of degree of damage, number of targets, risk of friendly fire, etc). The comparable decision point for a fighter is which target to attack. The prospect of missing rather than hitting - ie of dealing no damage - is, in my view, no more fundamental to the latter than the former.