At what point a list too large?

Hard to answer in such a fuzzy-set matter. I have no idea on the research and I would not believe in such research much, because the real problem is the criteria.

So, just a few notes.

A list size would be perceived mostly in comparison with the screen and ability to see the whole list at once. Even if this is somewhat irrational, people only really believe in data which they can observe at once. So, the answer will depend on screen size and font size, which is individual to each user but not dramatically different. Anything which is not visible on one screen without scrolling will be avoided by most users. In the cases when the list size is about the screen height, people would avoid applications which do not allow to expand the way to show the whole list. I guess, even if they almost never use this list. The application can be easily blacklisted if it happens just once.

There are exclusions where people will be irritated but will readily agree to tolerate a long list. In particular, it will happen if well-known cases such as list of countries, alphabetical. People will tolerate a long list when it's expected. Other factors include importance of information and the level of rationality. People will tolerate what they think is rational or just interesting, or hard-to-find.

However, the irritation with too long list will start from just more than 10 lines. Or 20, or 30. Here is the case where the individual preferences will be the most observable.

I think you have a misconception about the preferences of social groups. For example, the least tolerant to long lists are certainly scientists, and most tolerant are probably gamers. Real scientists like to think on conceptual level and simply do not perceive raw data like figures — outside some specific area. For example, most normal physicists are very good at estimation of sizes of objects in centimeters or light-years and time spans of many processes but will hardly agree to calculate a some of money to be spent on shopping; same thing with long lists; they will be reluctant to work with a list of more than 10 (also individual). Also, real scientists usually hate idiocy; and the long list is often a good indication of such "quality".

I think, with children the situation will be rather complex: they will agree to work with longer lists by might strongly over-estimate their ability to work with long list; so in real life they will make more mistakes; it depends on the age dramatically as well as individual traits; kids are not yet "polished" by society and keep more individual traits in simple tasks.

It would be interesting to see some experimental results and estimate my guesswork, but the experiments are hard to make correct. In my experience, I've been offered social some research questionnaires, but never was able to fill a single one in: the level of idiocy of questions was not acceptable to complete the task. :-)

—SA