Mr. Jack, Wounded King and others have responded--without reservation--that kin selection is responsible for the foul taste. So I gather that this is a widely accepted explanation. And kin selection is a form of natural selection.
I understand how kin selection works once the foul taste is established throughout a population, but it is harder to see how the foul-taste blood line gets started and spreads through a population by natural selection.
To simplify the matter, I first thought of a population of bugs that tasted okay. Then one individual bug in that population turns up with a genetic mutation that gave it all-out horrible taste, that a predator would reject. This is unrealistic, because the development of the foul taste would probably come in increments, rather that all at once.
At any rate, how does natural selection spread this gene from the first individual possesing it to the rest of the population? Might the gene come with a cost, and hamper the first individual in some way? Would it be able to attract a mate as easily?
But RAZD, you have the answer with the "neutral mutation." Neutral mutations can be used to fill in a lot of blanks, I would guess-- kind of a get-out-of-jail-free card. It doesn't seem to me that the spread of neutral mutations, in this case, can be considered natural selection. It is chance.
If the foul taste came about de novo incrementally, or if it existed for another purpose but was later refined for defense, it still seems that if neutral mutations were involved, the taste would be a result of chance more than natural selection.