Many people would admit that any given text, written in a natural language, might be further edited, polished, and refined, therefore this process could be, theoretically, endless (the possible little inferno for aspiring writers).
But when is it really reasonable for an author to stop?
I woudn`t like to sound teachy, since I`m writing here mostly about my experience.
I will try to take an unusual, “chemical” approach to this issue, concerning different types of texts, i.a. literary and also scientific (this – perhaps – to a lesser extent).
I think, the answer is rather simple: the right moment to stop with the editing process is when the number of new errors being made (also understood as clunky style and plot in some fragments) is comparable to the number of previously existing errors being deleted, that is when some kind of equilibrium has already been established – quite similarly to those chemical reactions which tend to run at the same velocity in the forward and backward directions.
Nevertheless, this idea is not so easy to apply in practice: the author who has come close to this very point, is usually so fixated on his/her editorial work, that it`s really hard for her/him to assess, how many errors he/she has already wiped out, and how many new mistakes have recently appeared in the content.
When I`m unable to see mistakes, I assume that the mentioned state of equilibrium is not far at all, and I usually send my text to at least one beta reader, then I put it away for two weeks or more (when possible), in order to read it just once or twice before submitting.
Polishing a text is most fruitful in the situation when efforts are, more or less, proportional to the results, that`s obvious. But this is often the case only at the beginning, and from a certain point, the more time you spend on rewriting, the less visible results you usually obtain.
This effect is quite similar to many others known from everyday life. For instance, a fountain pen which costs 200 dollars, would be probably of considerably better quality than one gained for 20 bucks. I expect, in turn, a model which costs 2000 dollars, to be just slightly better as a writing device than a pen for 200, and so on… That`s why it would be extravagant to spend a fortune on a pen when the author just needs one for a pleasant, smooth writing experience.
The efforts (e.g. money or work) invested into gaining something that could vary in quality, are in most cases proportional only to some degree to results expected.
Ultimately, it is up to authors to decide when the likely improvement in outcome would be rather miniscule in comparison with their further workload.
Proofreading: Monika Bajer: link to official site.