David Goodman wrote:
Having various reference techniques is very useful for people writing articles, who can choose whatever they feel comfortable with; having multiple simultaneous techniques is not quite as helpful for people trying to make small edits and fixes in articles, or adding references, because you need to be familiar with every individual one of them you might encounter. Personally, for example, I never use the cite templates if I'm adding refs to an unreferenced article, but i need to know them in case I work on an article already using them. And similarly with every possibility. I would rather have to learn any one thing, whether or not I dislike it, than need to learn them all. I recognize of course that this tends to inhibit experiment and improvement.
This is well taken. A lot of the templates have developed on an ad hoc basis, and when these become established there is a powerful unwillingness to change something that people are habituated to. With multilayered tranclusion it becomes even more difficult to adapt templates to circumstances.With large quantities of existing templates it may very well be that you have no way of knowing that the template that you need already exists.
Perhaps each group of templates needs a global review from time to time to see that the templates work together.
Ec