This is a question I have been asked many times, and have thought about myself.
Matthew Cobb over at Why Evolution Is True gave his take on this a few days ago
I basically agree with Matthew, but want to add something. I have two main points.
First, it is really hard to go "backwards", evolutionarily speaking (to invoke horrible teleology)- Matthew alludes to this as well
This is why a recent discovery of toothed frogs made everyone very excited.
Ordinarily when an organism "evolves" a new adaptation, or evolves "out" of an adaptation, they don't go backwards and re-evolve it. It is worth noting however, that it is not impossible for evolution to "go backwards", it is just unlikely.
Which leads me to my next thing, which is itself divided into two sub-points:
1. There really isn't any point to insects re-colonizing the ocean. Insects emerged onto land and found a colossal number of niches to occupy, and they are ridiculously successful. There was never any "reason" in an ecological sense for insects to go "back" to the ocean.
2. Insects are already in the ocean.
Given everything I just said, this second thing seems confusing, but it is related to the picture above.
You see, molecular phylogeny pretty much places insects within the same group as crustaceans. That is, insects are crustaceans, in the same sense that birds are dinosaurs. Insects evolved from (some) crustaceans, just like birds evolved from (some) dinosaurs. The picture accompanying this post is a remipede, and is currently considered the "sister" taxa of insects- that is, the remipede is to insects what the chimpanzee is to us humans.
Take a look at this huge phylogeny tree! Insects (Hexapoda) are at the top, and the closest related group to them is the remipedes (Xenocarida).
Interesting to note, and something I won't go into, is that tardigrades and onycophorans are considered "outgroups" here. Previously these were considered to be the "sister" groups to insects (the chimp/human analogy)...but no longer, it seems.
So, in short, there just isn't any point for insects to go "back" to the ocean, especially when they are essentially already there.
(yes, some of this research is a year old, but I just found it, and thus it is new to me!)
(Citation for above picture, phylogeny tree, information, etc)
Regier, J. C. J. W. Shultz, A. Zwick, A. Hussey, B. Ball, R. Wetzer, J. W. Martin and C. W. Cunningham. 2010. Arthropod relationships revealed by phylogenomic analysis of nuclear protein-coding sequences. Nature 463:1079-1083.