No magic bullet. Here's the process:

--All feedback gets pushed to my personal email inbox--gmail :)

--I try to do 0-inbox, i.e. keep my inbox at zero messages. This is usually not attainable, but means it functions essentially as a to-do list.

--I respond to all feedback ASAP (unless anonymous).

--If it is something simple that I know how to fix I try to do it that day.

--If it is something a bit more complicated, I respond, and put it on the bug queue. I'm currently using http://speckleapp.com for that, made by an HNer (http://elliottkember.com/).

--Every few days I set aside a large block of time to go through that bug queue. If easily fixable, I fix it and respond back: "Fixed!" If not, I put it in another category, and explain why it is a complex issue and how and when it might get fixed (or not).

There is so much that can still be done as computational linguistics makes its way into how search queries are understood and how content is indexed and retrieved. Faster machines will make some things possible that were not (like instant search) soon!