The best answer to this question that I've heard was given by Maddrell: avoidance of visual predation. Zooplankton avoids it by diurnal migration ("Diel vertical migration"): every morning it dives down below 1000 m into the dark waters; during the night it feeds in the upper column of warm water. The advantages of this energy-consuming strategy are so significant that they outweigh disadvantages:
Unlike the crustaceans in zooplankton, insects do not have proper gills, they have trachea and air sacs. The problem is that these fail under high pressure.
...In the open sea, the only escape routes from fish predation that might be available to insects are likely to be those used by crustaceans. Crustaceans make extensive diurnal vertical migrations that remove them from the upper reaches of the sea in the daytime when they are visible and so at risk. Insects cannot adopt the same strategy because, for virtually all of them, their tracheal method of respiration depends on the presence of gaseous air in at least parts of the body. At 100 m depth, for example, the total pressure is above 10 atm, and this would so reduce the volume of air in the system that it fails. In addition, the use of a surface film of air as a physical gill, the plastron, by which many freshwater insects can live for prolonged periods under water without requiring access to the surface, fails at depths below approximately 30–40m because the pressure causes bending of the supporting hairs and the collapse of the air film. So insects could only survive in the sea if they stayed in the depth range of the top few tens of metres. Unfortunately for them, there is no escape there from daytime predation by fish. They cannot, for example, reduce predation by becoming transparent, an adaptation of some small crustaceans. That would merely reveal their glistening air-filled tracheal system which is so obvious in any insect dissection. Neither could they become sufficiently unpalatable to ward off predators. If this were possible, then the same adaptation would have been selected for in crustaceans. http://jeb.biologists.org/content/201/1
Maddrell gives an interesting example of a race between predatory fish and midge larvae in anoxic African lakes:
...larvae of Chaoborus avoids predation by diving down to 70 m. It can avoid the limitation that its tracheal system must fail at these depths from the fact that it is reduced to two pairs of gas-filled spherical sacs used to provide variable buoyancy. It survives well in Lake Barombi Mbo (West Cameroon), although there is essentially no oxygen below 20 m. The cichlid fish that is the major predator of Chaoborus, has blood with an extraordinarily high haemoglobin content, presumably to allow it better to prey on Chaoborus larvae in anoxic water as these dive at dawn and ascend at dusk. It seems that larval Chaoborus are uniquely adapted to avoid asphyxiation by using an anaerobic malate cycle to generate ATP.In addition, they are able to migrate vertically using a buoyancy control mechanism in which the cuticular walls of the tracheal sacs are made to swell or shrink.
Diving down to 70 m is an astonishing feat, but at this depth the visibility is still sufficient to prey on larva: it avoids predation only by residing in anoxic waters that asphyxiate fish. Furthermore, there is an enormous difference in pressure between 70 m and 1000-1500 m. The lesson is that in the oxic ocean, insects cannot outcompete migrating crustaceans.
The situation would be different in anoxic oceans. Many times over in Earth's history there were long periods of anoxia in the ocean that led to marine mass extinctions. Not only such extinctions clear the oceans of predators; these episodes give advantage to air-breathing divers grazing on phytoplankton under no pressures of predation and competition. It is likely that insects had multiple windows of opportunity, but the subsequent oxygenation evry time robbed them of their chance to establish themselves in the open ocean properly. The most successful land animals have been denied re-entry to their ancestral habitat! This is an outrage. Speak of injustice of nature...
But there is a way out of these wrongs. At present, the most abundant marine insects are water striders that live on floatsam; they've colonized open seas many times over the last 45 Myr.
The problem with this strategy is that the debris (rafts of vegetation, algae, pumice, tar, feathers, etc.) are infrequent and tend to sink rapidly, which makes it difficult to make it out big in the open ocean. Here comes the long-awaited opportunity: our plastic trash (that disintegrates to virtually unsinkable pieces of a few mm across that are perfect for the bugs) is the new habitat for the striders. We have already filled with garbage significant part of the North Pacific Subtropical Gyre that covers an expanse the size of Texas - right in the middle of nowhere. The population of water striders is exploding over there.
Finally, after millions of years of evolution, insects have the real chance of making it into the (oxic) high oceans: something that eluded them forever.
Isn't that good news?