I think I understand your point, that one shouldn’t create unneeded features. However I think there are someways in which this model doesn’t quite fit.
First, we don’t have a way to collect data from the people who gave up on using the site because they had problems. That I can think of. That makes it hard to do decision making in the way you propose. So part of the idea with the project is that it would be tried as an experiment. If after 3 or six months it proved to be unnecessary, or if we gather enough data on the problems people have so that we can solve them some other way, then great. We’ve learned something. And we’ve improved non-live help or reduced the major issues.
I think it’s also important to remember that what I imagine as live chat is not part of the suttacentral.net website. There would simply be a link to a help website from which people could do a chat if they wanted. So the danger of it complicating the main website would be near zero, eh?
Oh, without a doubt there is no need for chat support for the majority of the users. The site isn’t that bad. So again, I’m having a hard time applying your method to the problem.
I think we can safely say that the majority of users who have a problem using the site would like to solve the problem they are having. So in that sense, we could say “majority of users with problems.” Then of course you would need to break that down into percentage of those users who would want chat to solve their problem. How to know that? If we had a help site, we could track usage and see at what point people opted for chat I suppose.
Personally, I’m coming at this from a Dhamma perspective. If I had the opportunity to help someone overcome the hurdles they are facing to access the suttas, then I would want to take it. Volume wouldn’t matter. And frankly the method of giving help doesn’t matter either.