Your questions go to the heart of a number of important issues associated with the evolution of social behavior; one could easily write a book in an attempt to answer all of your questions adequately (in fact here is a technical book on the topic: Buss 1994).   Let me try to provide a good “introductory” answer.

We will start at the simplest biological level. A number of observable facts and their logical implications lead to the theory of evolution. When we are talking about simple, asexual unicellular organisms, the process is relatively simple, but things start getting (increasingly) complicated when organisms cooperate in order to reproduce (and sexual reproduction is the most common form of reproductive cooperation).

An obvious question is, why would organisms ever cooperate? One might think that cooperation would be counter productive compared to selfish (asexual) behavior, at least from the perspective of producing the maximum numbers of offspring per organism. On the other hand, sexual reproduction has clear benefits in terms of genetic and phenotypic diversity [see Otto, 2008].

Cooperation, sexual and otherwise, can have many forms and has both costs and benefits. What is important evolutionarily is that for any type of social behavior (like sex) to evolve, the reproductive benefits must outweigh the costs (and here the issue is complicated by the need to considered the relatedness of individuals - for example, if you sacrifice your life to save (and help insure) the reproductive succcess of your genetically identical twin, then that type of altruistic (self-sacrificing) behavior can be selected for.

This is a phenomena known as “inclusive fitness [see Dugatkin, 2007]. The situation is complicated, however, because a population that supports social behavior will also have to police itself against “selfish cheaters”, that is those who attempt to capture the benefits of social behavior without paying the price.  That is a large topic on its own, but related to the questions you are asking. For example, cancer is a type of (ultimately self-destructive) anti-social behavior and multicellular organisms have evolved mechanisms to suppress cancer. 

Now consider the role of love (and bonding) in human reproduction (similar reasoning applies to many different types of animals.)  What features of human behavior are linked to, and dependent upon, social cooperation?  Beyond the sexual act itself (a form of cooperation), human babies are born helpless (semi- premature), and they are completely dependent upon their mothers for food and protection. This has a benefit, human brains are born under-developed and grow dramatically post-natally).

Human pregnancy renders females more vulnerable, so there is a benefit to recruiting the help of others. In humans, this generally involves the formation of stable mating pairs, as well as larger social groups (think communities, tribes, cities, and nations).

Clearly (and particularly in humans), there is a large cost to a female who gets pregnant through a mating with an “unreliable or unhelpful” male (and there are biological processes associated with whether pregnancy occurs or pregnancy is carried to term (Knapp & Innocent, 2012).

A female might “prefer” to select the most robust, but not necessarily the most loyal, mate while at the same time securing the help of a “loyal partner”. Similarly, a male who commits to the support of a single female both restricts the number of potential offspring they could father, but dramatically increases the odds that the offspring they do produce will survive and prosper. That said, it becomes critical to the male that the offspring that are produced by “their” female are genetically related to them. In some species this leads to "mate guarding" (a form of jealousy). 

So we are talking about the effects on reproductive success (the only outcome biological processes “care” about) of love, bonding, jealousy, loyalty, etc. These traits help (but certainly do not guarantee) reproductive success.

There is also a conflict between being in love and being able to fake being in love (for example).  This means that it is a valuable trait to be able to detect fakes. The reciprocal process of falling in love can be seen as a way to test the commitment and trustworthyness of a potential partner. 

Many human traits are involved in sensing what others are thinking. Those who cannot sense emotions (or whose emotions cannot be read) are likely to be at a reproductive disadvantage in a species, like humans, that depend on “mating rituals”, such as “falling in love”.

This is yet another gigantic topic - social signaling and the biological basis of “intuitions”, feelings, and social behavior.   As an example, feelings of pain are often associated with social rejection, and these feelings involve the same regions of the brain that are associated with physical pain (see Eisenberger et al. 2003).

You are right that there are possible alternatives to the human-specific strategy.  For example naked mole rates represent an extreme form of communal behavior (Burland et al., 2002 and link) in which only one female is reproductively active (the opposite of polygyny).

But if you think about the biology of the human species, you can probably give a good explanation for why there are no (extant) human societies based on casual sex without bonding or “punishment” / ostracism for failure to bond or to act in a socially acceptable manner.

That said, it is important to realize that biological processes while effective evolutionarily, are not necessarily humane with respect to individuals.

Whether humans can transcend their biological roots and build better (less painful, but perhaps less rewarding) social interactions, including love remains to be seen.