This is all very well thought out, but really isn’t addressing my use pattern. perhaps it works for others.
I find that I don’t use feedly’s feed search at all: I add feeds directly from the RSS icon on a page that I’ve navigated to through the usual link-following process and think that I’ll track it for a while.
Much more useful to me would be a process that tracked the fraction of articles from each feed that I actually open to read beyond the headline scan. That way I could deprecate and eventually un-follow the feeds that turn out not to be as useful as they first seemed.