It is hard and frustrating to work on a revision and a letter to the referee, in general, but particularly when the editorial board finds it hard to decide whether to ask for a revision or that the paper is not suited for their journal.
In weak moments, I see the value and purpose of the peer-review process the way it currently works in economics and most other scientific disiplines today. They are, however, weak moments. What comes out of the peer-review process? Usually a long list of minor issues which, for the general finding, has no real importance. Sometimes, of course, major flaws are pointed out. Ultimately, peer-review (is supposed to) guarantee quality and relevance.
How would the world be like without the process? Crazy, perhaps? Would it be a world where one could not trust the written word, and where quality, relevance, and importance were without meaning? Of course not. Instead, every journal would have an open-source, continually ongoing review process, where the responsibility for quality and relevance was placed solely with the author and the editorial board; where poor work would be openly critisized in responses and comments; where authors could focus more on developing ideas and writing skills (instead of excuses, irrelevant details, and cover-up operations); and where authors would be forced to think harder about problems before submission (and not during the revision). Scientists would perhaps write more monographs and fewer articles as the main difference (the peer-review process) ceased to exist, and as the monograph are better suited to report on many scientific findings. Citations would be a better measure of importance and influence; today, a good deal of the references are included on the whim of referees. Finally, perhaps it would dawn on us that knowledge evolve, and that to be knowledgable on a given subject requires familiarity with a whole literature, not only a handful of articles from the leading journals.