Incrementalism is what makes evolution credible. One mutation and a resulting change in survivability, followed by another... So is it not possible that this fish got to its present state through small changes? I think so.
Fish have had hundreds of millions of years to learn how to capture insects on or near the water. Reasonably, strategies that made sense and that were genetically selected involved getting the prey wet. From getting insects wet, to spitting on them, to spitting accurately, is very reasonable chain of small changes.
If you're basing evolution on 'advantageous' mutations, the theory fails. There are no such examples of beneficial mutations.
Plus, mutations do not create any 'new' information, it only is able to affect whatever dna information is already there.
Mutation works such as this: let's use an analogy here and treat a word as a piece of genetic code. Say the word is "Christmas".
A cosmic ray hits the word and changes something. You can get "Christ", or Chris, or mast, or rats, but you can never get 'food', book, menu, emu, etc.
Same with the mutations that are observed in nature. You get five-legged cows, or if you've seen the most recent example I've seen, a baby with three arms. The extra arms are not functional, and don't beneficially add to the organism. Same with two-headed turtles or snakes. You will never see a cow with a beak, or wings, because mutations only affect the dna code the animal has.
Besides, dna's main function is to PRESERVE the life-form from CHANGING into something else, not so that x-rays can blast it and scramble it into something else.