Something like 30 out of the 500 anopheles species are capable of carrying malaria. Only 2 of those 30 bite humans:
A. gambiae and
A. funestus. Eradicating those two would be sufficient and another species, one that doesn't bite humans, would take over the rest of their niche. "Bites humans" is not a sufficiently beneficial adaptation to be a game-changer. We can eliminate that trait and
precisely that trait with surprisingly little selective pressure.
Most of the "whoops" tales we hear are people that didn't listen to their scientists telling them not to, didn't have scientists at all, or were from long enough ago that we can discard their stories as irrelevant in the same way that we'd discard things like "spaceflight is impossible" and "nukes will end war". We've now traced the food chain around malaria mosquitoes out to three or four degrees of separation, including things like their competitors, the diseases they carry, and the things they spread those diseases to. Their entire ecological niche is trivially replaced by close relatives that don't carry malaria to humans. They're not even common enough to have major food-chain or crowding effects like you'd see with ants, rabbits, kudzu, or sparrows. Everything that we can find says that the damage from eradicating these two species would be zero. Not "small", not "manageable". None.
Malaria-carrying human-biting mosquitoes are small enough and unimportant enough that we should be thinking of them as disease organisms rather than insects. I'm sure you're perfectly happy with the eradication of bot flies, polio, and smallpox; why not A. gambiae and A. funestus?