Most of my research leads to 20-60 page papers, which need to go through a refereeing process in order to appear in print. I have received positive, negative, and indifferent reports.
The positive ones are the nicest to receive, and they often contain lots of useful suggestions for improving the paper.
The negative ones hurt the most. I usually cannot read them the entire way through without shaking. Negative reports are generally are one of two types. Either the referee does not like the paper’s idea and also complains about the implementation. Or, they think that the idea has some promise, but argue that the implementation would require too much effort to improve. Both types are helpful, if the reports are not wishy-washy. I can fix the implementation, if I think that the idea is good enough. Then it’s time to try another journal. I have often found the negative reports the most helpful—for that paper and for improving my future papers.
But it’s sometimes a game. I now try hard to write the paper in a way that they only can complain about the idea, not the implementation. Impossible, but I try.
The worst reports are the indifferent ones. They usually are rejections, and that means that the reviewer did not like the paper. Fine. Indifferent reports often don’t help me move ahead on improving the paper, though. So then its just a quick polish and off to another journal to play the roulette wheel again.
My overall impression is that refereeing helps the paper—if I can figure out what the referees did and did not like. But referees may not always be truthful. Or the letter to the editor may be different than the referee report.