It's worth considering that from the cellular point of view, cancer is extremely 'successful.' But at the human level, discovering a tumor is not a cause for celebration, but for anxiety and further investigation. Some tumors are benign, and can be thought of as a mere excess of a good thing causing only cosmetic or minor discomfort. Others are malignant, actively reproducing themselves to excess and pulling all the body's resources towards themselves, disregarding the health of the host body but without creating any sort of value (because of the impossibility of reproducing itself independently of its originating context). A benign cancer is tolerable; a malignant one must be excised or it will kill the host.
(This might seem like a hand-wavey objection, but systems biology suggests that Pareto effects show up in the development of cancers just as in many other natural systems, eg http://journals.plos.org/plosone/article?id=10.1371/journal.... )
We should not fall into the trap of assuming that just because a social phenomenon instantiates some natural behavior, that it is therefore a Good Thing, aka the naturalistic fallacy. Nature is inherently fractal, and extreme outcomes may be instances of local maxima that are suboptimal from the perspective of the system as a whole.