I will try again, with special attention on not using any buzz words. The OP asks a specific question, why has a trait not developed, since that trait would obviously benefit the individual bee, and should therefore be conserved, any time it occurred. This assumes a set of conditions exist for what should happen in evolution.
But the conditions that actually exist are that individual survival is not as great a factor, by many orders of magnitude, than the successful creation of a new generation of the species. The bee in question is overwhelmingly likely to be a worker, and even more to the point a worker in the later stages of it’s life. (More mature bees predominate among the likely guard contingent of a hive.) This class of bees will have no progeny of their own, however long their lives may be. Their deaths have no effect of species survival. Preservation of their lives provides no benefit, and their deaths make resources available to the next generation.
Let us examine a different case. American White Tail Deer in an undisturbed environment will expand their population to the limits of predation, and food supply. This process involves an annual death rate from starvation, and predation on the order of one third of the herd. However, the distribution of such deaths is disproportionate for the sexes. Female deer are more timid, and have a more efficient metabolic use of their body fat than males. The combination increases death by both predation, and starvation among males. As many as half the males in a healthy herd may die every year.
Some of the reasons for this are that males have much higher activity levels, higher muscle tone, antlers, and competition with other males for mates that causes them to spend their energy at a far greater rate than the females. In addition, males engage in behaviors that make them preferentially available to predators over females. None of these characteristics increase the survivability of the individual male deer, but they are conserved. In fact, female deer preferentially select mates based on the higher level of such characteristics, such as antler size, and herd dominance. If that is true, it must therefore be true that female deer that do so have more surviving descendents than females who do not.
Death in higher numbers is a survival characteristic for the herd, if those higher numbers include more dead males. Dying is an adaptation for species survival. If more males die, more females survive. Since one male is predominately responsibility for the genetic contribution to the next generation for a large percent of the herd, a male with these death enhancing characteristics is more successful in producing reproductions of his own genetic code. His daughters will carry the genetic pattern that favors mate selection on that criteria, and his sons will have more of those characteristics that he carries.
The existence of socialization among animals profoundly changes what characteristics are beneficial to the species. The so called altruistic behaviors often debated are sometimes confused with the existence of a decision on the part of the individual animals. If I wrongly attributed that to the OP, I apologize.
Tris