Yes, John's work is seminal, and fair (flow) queuing is essential in cleaning up the bufferbloat mess. Flow queuing has a bigger effect on end user experience than mark/drop algorithms to signal TCP (though that is also needed to keep TCP from losing its mind). I distinguish flow queuing from fair queuing as it has aspects that are not "fair", but the fundamentals are the same.

The recent BBR work at Google is another key to solving bufferbloat.

I finally think we have the tools needed to really solve the problem, between flow queuing and new mark drop agorithms (e.g. Codel, CAKE, PIE, fq_codel, etc.) and BBR.

So again thanks to John for his key work.

Jim Gettys (who coined the word bufferbloat).