Let’s be utilitaristic and do a rough, optimistic calculation:
First, how much time does correcting a spelling mistake cost?
- First of all, it costs you some time to find a way to contact somebody from the journal. I ran an experiment with a random journal and it took me three minutes to find a way to contact the chief editor and to ensure that there is no easy way to contact somebody closer to actual typesetting.
- It costs you about one minute to write the mail.
- It costs the chief editor (in our example) at least one minute to read the mail and redirect it to typesetting.
- It costs the head of typesetting at least one minute to delegate the work.
- Whoever is actually doing the work, has to find the source of the paper, correct the mistake, check whether the whole paper is still neatly arranged (even if the correction did not alter the number of lines in the paragraph, the linebreaking algorithm or the font might have slightly changed) and every sentence is still on the same page (in fields, where references to pages happen). Again, I ran an experiment on one of my own papers and it took me two minutes (as the correction had no major effects) – and I operate my computer mainly via keyboard and consider myself a fast typer and well organized, not to forget that I do not have so many papers. Additionally, let’s consider one minute for uploading or similar.
So, all in all, mankind has spent nine minutes on correcting the mistake.
Now, At the end of the day, the reason why we bother with spelling is that it speeds up reading texts, i.e., saves the reader some time. Let’s say a small spelling mistake (like turbulance) costs every reader a second. Thus to break even with our nine minutes, the respective word has to be read about 500 times. I do not have any direct numbers on this, but as papers are mainly read by the same people who write papers, this would mean that I have to read (every word of) 500 papers for every paper I write – which is very far from reality.
Thus even under the above optimistic conditions you are likely wasting more time than you are saving.
Also, from another point of view and my own experience: The involved people do not care. I can tell a story of how much trouble it was to make a typesetter change the way one of my tables was formatted such that it was any readable even if that would make it deviate from the journal guidelines – and that was a severe issue.
Finally, it’s likely that you only need to read a few papers to spot a terminus technicus with confusingly wrong hyphenation. My favorite example is »generalized onset seizure«, which denotes seizures with a generalised onset and not onset seizures, which are generalised.