It's not that easy. We have earlier had more case insensitive searches, but there were so many conflicts all over the place.
For standard A-Z it's quite easy. But when you mix in UTF8 it's quite hard. Which of these should count as the same character?
E, e, é, è, É, ë, ê and so on.
We do however see your point and might get back to it and make active design decisions here and there. But as I said, it's not easy at all.
The behavior is also dependent on which search type you do. If you do an extended search it will be case insensitive in most cases. Example: https://goo.gl/GQiYL6