After reviewing the arguments of species as real entities vs. species as made up configurations for the past several hours, I remain a pluralist on the topic of species concepts.
The Biological Species concept by Ernst Mayr (members of metapopulations that actually or potentially interbreed) works quite well for large animals that undergo only sexual reproduction. Such organisms can be easily observed for discrete interbreeding over time. But what about those populations that show discrete yet sympatric behaviors and habits in the field yet interbreed easily in the laboratory. Would these be considered species? Do these actually offer discrete units? Under the Ecological Species concept (isolated phenetic clusters of individuals sharing a particular niche) they would.
This is all well and good for laboratory animals. But how about those organisms that don't easily take well to experimentation? How about asexual animals, where breeding doesn't occur, or those organisms (like plants) that easily hybridize? Do we once again scrap "species"?
And yet, there are other concepts we can use (a good list is in the book Speciation - Coyn and Orr) that do not rely on a strict and exacting understanding of the direct mechanisms or processes to obtain understanding. In systematics, we aim to describe and understand biodiversity. The title "species" is used in multiple ways to best describe different organisms. Of course we would love to test the biology and ecology of every candidate species on this planet; almost always this is currently impossible or unfeasible. Unless we are to simply give up on describing and understanding biodiversity, there needs to be some way to separate and describe units we believe are candidates species. With the biodiversity crisis, we cannot wait for every or even most units to be verified. Operational concepts provide a framework for describing species that the Biological Species concept could not. Homeostatic Property Cluster theory provides basis for species in asexual organisms like Eubacteria and Archaea.
So I remain a pluralist. A scientific method should make the world LESS confusing and not more so. Rejecting the reality of species or retaining only one species concept means that only very little of the diversity on this planet will make sense. And even if species are just arbitrary human conceptualizations making discrete compartmentalizations of continuous diversity, well, how else would you suggest making any sense of it? A good scientist knows that the best way to tackle a problem is to break it down into workable pieces.